Zainab's Report SRS

INFORMATICS INSTITUTE OF TECHNOLOGY
Department of Computing
(B.Eng.) in Software Engineering
5COSC021C2: Software Development Group Project
Project SupAut: Teaching Using Effective Assistive Technology To Support

Language And Communication Development In Middle School Autistic
Students
Module Leader: Mr. Banuka Athuraliya

This project is partial fulfillment of the requirements for the Software Development Group Project
Team: SEMICOLON
IIT ID UoW ID Student Name
20210482 w1867602 Mohamed Fahim Shafna Zainab
20200730 w1833551 Munasinghe Arachchige Sandil Methmin Munasinghe
20210443 w1870513 Muhammed Azahim Muhammed Adheeb
20210729 w19128867 Mallawa Witharanage Minoli Kamwinie Dayarathne
20210129 w1870597 Melewwe Thanthrige Mewan Manodhya Amarasinghe

Declaration
DECLARATION
We hereby declare that the content of this project report and all related artifacts are original work and
have not been previously submitted or are currently being submitted for any academic program.
Student Full Name : Mohamed Fahim Shafna Zainab
Registration No : 20210482 | w1867602
Signature :
Student Full Name : Munasinghe Arachchige Sandil Methmin Munasinghe
Signature :
Student Full Name : Muhammed Azahim Muhammed Adheeb
Signature :
Student Full Name : Mallawa Witharanage Minoli Kamwinie Dayarathne
Signature :
Student Full Name : Melewwe Thanthrige Henry Mewan Manodhya Amarasinghe
Signature :
Date : 13th January 2023
SEMICOLON - SE – 62 |i
Declaration
ABSTRACT
This research study aims to investigate and identify effective strategies and tools for tracking and
documenting the progress of autistic students in traditional inclusive classroom settings. The primary
objective of the study is to address the challenges faced by educators in terms of enhancing the social,
creative, and communication skills of autistic students. A mixed-methods approach was employed,
including literature review, case studies, interviews, and surveys, to gather comprehensive data on
the most pressing challenges faced by educators in teaching autistic students. This approach enabled
the research team to gain a deeper understanding of the situation and identify the key areas of concern.
Subsequently, a variety of research, design, development, and evaluation methodologies were
implemented to filter the most efficient solutions. These solutions were developed and implemented
with the goal of supporting educators in their efforts to enhance the social, creative, and
communication skills of autistic students. Furthermore, the study explores the potential of assistive
technologies, such as conversational AI, AI art generators, NLP techniques, and transformer-based
language models, to support the language and communication development of 7-12-year-old autistic
children. These solutions are developed to improve the classroom experience for the students and
teachers alike. The findings of this study furnish insights on ways to systematically monitor and
document the progress of autistic students and bolster their language and communication skills
through assistive technologies in traditional classroom settings, thus bridging the gap in current
research in this area.
Keywords: Autism, Inclusive Education, Teaching, Assistive Technology, Social, Creative,

Communication Skills, Language and Communication development, conversational AI, AI art
generator, NLP, Transformer-based Language model.
SEMICOLON - SE – 62 | ii
Declaration
ACKNOWLEDGEMENT
We extend our deep appreciation to Mr. Kushan Bhareti, our mentor, and Mr. Banu Athuraliya, the
leader of the Software Development Group, for their guidance and support throughout this project.
We also extend our gratitude to the other instructors who provided invaluable instruction on how to
compose this report. Our thanks also goes to our classmates who participated in the questionnaire and
offered assistance in various ways, as well as those who provided feedback on the report's
composition. Our team members deserve special recognition for their diligent efforts in putting
together this report. Finally, we are grateful to our friends, family, and loved ones for their support
throughout this endeavour.
SEMICOLON - SE – 62 | iii
Declaration
TABLE OF CONTENT
DECLARATION i
ABSTRACT ii
ACKNOWLEDGEMENT iii
TABLE OF CONTENT iv
LIST OF FIGURES xvi
LIST OF TABLES xvii
LIST OF ABBREVIATIONS xix
CHAPTER 1: INTRODUCTION 1
1.1 Chapter Overview 1
1.2 Problem Background 1
1.2.1 Autism And Inclusive Education 1
1.2.2 Teaching in Inclusive Education 2
1.2.3 Technologies to Support Teaching in Inclusive Education 3
1.3 Problem Statement 5
1.4 Research Gap 5
1.5 Research Questions 5
1.6 Research Aim 6
1.7 Project Scope 7
1.7.1 In-scope 7
1.7.2 Out-scope 7
1.8 Rich Picture Diagram 8
1.9 Objectives 8
1.9.1 Research Objectives 8
1.9.2 Academic Objectives 9
1.9.3 Operational Objectives 10
1.10 Resource Requirements 11
1.10.1 Hardware Requirements 11
1.10.2 Software requirements 11
1.10.3 Technology Stack 13
1.11 Chapter Summary 13
CHAPTER 2: LITERATURE REVIEW 15
SEMICOLON - SE - 62 | iv
Declaration
2.2 Concept Graph 15

2.3 Domain 15
2.3.1 Autism 15
2.3.1.1 Prevalence 15
2.3.1.2 Diagnostic Criteria 16
2.3.2 The Inclusive Education 17
2.3.2.1 Integration and Mainstreaming 17
2.3.2.2 Inclusive Education in Sri Lanka 17
2.3.2.2.1 Challenges to the Implementation of Inclusive Education in Sri Lanka 18
2.3.2.3 Approaches to Inclusive Education 19
2.3.2.4 Inclusion for Students with Autism 19
2.3.3 Teaching in Inclusive Education 20
2.3.3.1 Challenges in teaching in an Inclusive Classroom 20
2.3.3.2 School Resources in an Inclusive Classroom 21
2.4 Technologies 22
2.4.1 Selection of AI Art Generator 22
2.4.1.1 Choosing the Right Conversational AI for Project SupAut 22
2.4.1.2 DALL-E- Under the Hood 23
2.4.2 Selection of Conversational AI 24
2.4.2.1 Choosing the Right Conversational AI for Project SupAut 25
2.4.2.2 Twilio - Under the Hood 26
2.4.3 Selection of Transformer-based Language Model 27
2.4.3.1 Transformer-based Language Model - Under the hood 27
2.4.3.2 Choosing the Right Transformer-based Language Model for Project SupAut 29
2.4.4 Selection of Natural Language Processing (NLP) Technique 30
2.4.4.1 Importance of Sentiment Analysis in the Domain 30
2.4.4.2 Role of Treebank Annotation in Sentiment Analysis 30
2.4.4.3 Choosing the Right NLP Technique for Project SupAut 32
2.5 Existing Systems 33
2.5.1 Importance of Technological Solutions in the Domain 33
2.5.2 Lack of Technological Solutions in the Domain 34
2.5.3 Evaluation of Eolutions in the Domain 35
2.5.3.1 Frontline Education 35
2.5.3.2 Cough Drop 35
2.5.3.3 Uniqueness of the Proposed Solution 36
SEMICOLON - SE - 62 |v
Declaration
CHAPTER 3: METHODOLOGY 38
3.1. Chapter Overview 38
3.2. Research Methodology 38
3.2.1 Research Philosophy 38
3.2.2 Research Methodologies 40
3.2.2.1 Quantitative Research 40
3.2.2.2 Qualitative Research 40
3.2.2.3 Mixed-methods Research 41
3.2.2.4 Comparison between Research Methods 41
3.3. Development Methodology 42
3.4 Design Methodology 43
3.5 Evaluation Methodology 44
3.5. Project Management Methodology 46
3.6 Team Work Breakdown Structure (WBS) 47
3.7 Gantt Chart 47
3.8 Usage Of Project Management And Collaboration Software In The Project 47
3.8.1 ClickUp 47
3.8.2 Slack 48
3.8.3 Google Meet 49
3.9 Risks and Mitigation 50
CHAPTER 4: SYSTEM REQUIREMENTS SPECIFICATION (SRS) 53
4.2 Stakeholder Analysis 53
4.2.1 Onion Model 53
4.2.2 Stakeholder Descriptions 54
4.3 Selection of Requirement Elicitation Techniques/Methods 55
4.3.1 Literature Review 55
4.3.2 Questionnaire 56
4.3.3 Interviews 57
4.3.4 Brainstorming 58
4.3.5 Prototyping 59
4.4 Discussion/ Analysis of Results 59
4.4.1 Questionnaire Findings 59
4.4.2 Interview Findings 59
4.4.3 Literature Review Findings 60
SEMICOLON - SE - 62 | vi
Declaration
4.4.4 Summary of Findings 60

4.5 Context Diagram 61
4.6 Use Case Diagram 61
4.6.1 Use Case Descriptions 62
4.7 Domain Model 70
4.8 Functional Requirements (with prioritization) 70
4.9 Non-Functional Requirements 72
CHAPTER 5: SOCIAL, LEGAL, ETHICAL AND PROFESSIONAL ISSUES 74
5.2. SLEP Issues and Mitigation 74
5.2.1 Social Issues 74
5.2.2 Legal Issues 74
5.2.3 Ethical Issues 75
5.2.4 Professional Issues 76
CHAPTER 6: SYSTEM ARCHITECTURE & DESIGN 77
6.2 System Architecture Design 77
6.3 System Design 78
6.3.1 Class Diagram 78
6.3.2 Sequence Diagram 79
6.3.2.1 Sequence diagram for the dashboard 79
6.3.2.2 Sequence diagram for the Chatbot 80
6.3.3 UI Design 81
6.3.3.1 Dashboard page 81
6.3.3.2 Add student page 81
6.3.3.3 Add question page 82
6.3.3.4 Student Details page 83
6.3.3.5 Question page 84
6.3.4 Process Flow Chart 85
6.4. Chapter Summary 85
REFERENCES I
APPENDIX XIV
Appendix - Section A - Literature Review XIV
Appendix Section A.1 - Concept Graph XIV
SEMICOLON - SE - 62 | vii
Declaration
Appendix - Section B - Methodologies XV

Appendix Section B.1 - Work breakdown Structure XV
Appendix Section B.2 - Gantt Chart XVI
Appendix Section B.3 - Project Management Methodology XVII
Appendix Section B.3.1 - ClickUP XVII
Appendix Section B.3.2 - Slack XVIII
Appendix Section B.3.3 - Google Calander - Scheduled Google Meet meetings XVIII
Appendix Section B.3.4 - Table of Meeting minutes XIX
Appendix - Section C - SRS XXI
Appendix - Section C.1 - Questionnaire XXI
Appendix - Section C.2 - Questionnaire Findings XXIII
Appendix - Section C.3 - Interview Findings XXVIII
C.3.1 Interview 1 with a SEN Teacher (2022) XXVIII
C.3.2 Interview 2 with a Trainee SEN Teacher (2022) XXX
C.3.3 Interview 3 with a SEN Recruiter (2022) XXXI
Appendix - Section C.4 - Literature Review Findings XXXIII
Appendix - Section C.4 - Use case descriptions XXXIV
Appendix - Section D - Design XXXVII
Appendix - Section E - Work Breakdown Matrix XXXVIII
SEMICOLON - SE - 62 | viii
Declaration
LIST OF FIGURES
Figure 1 : Rich Picture Diagram of Project SupAut (Self - Composed)

Figure 2 : DALL-E 2 image-generation process (Ramesh, A. et al, 2022)
Figure 3: A diagram that illustrates the Diffusion Model (Ramesh, A. et al, 2022)
Figure 4: Initiating an outbound message and handle task creation (Twilio, 2022)
Figure 5: The encoder-decoder structure (Vaswani et al., 2017)
Figure 6 : Dependency Based Annotation (Rajput, A. 2019)
Figure 7: Constituency Based Annotation (Rajput, A. 2019)
Figure 8 : Team Work Breakdown Structure
Figure 9 : Gantt Chart Diagram
Figure 10 : Task breakdowns for each group member in ClickUp.
Figure 11 : Snip of Gantt chart usage in ClickUp.
Figure 12 : Slack application for group collaboration and effective communication
Figure 13 : Snapshot of a Google Meet session
Figure 14 : Snip of a video conference using Google Meet
Figure 15 : The Onion Model (Self Composed)
Figure 16 : The Context Diagram (Self Composed)
Figure 17 : Use case diagram(Self Composed)
Figure 18 : The Domain Model diagram (Self Composed)
Figure 19 : The High Level Architecture Diagram (Self Composed)
Figure 20 : The Class Diagram (Self Composed)
Figure 21 : Sequence diagram for the dashboard(Self Composed)
Figure 22 : Sequence diagram for the Chatbot(Self Composed)
Figure 23 : The Dashboard page
Figure 24 : The Add student page
Figure 25 : The Add question page
Figure 26 : The student details page
Figure 27 : The Question page
Figure 28 : Concept Graph(Self Composed)
Figure 29 : Section breakdowns in ClickUp
Figure 30 :Task Checklist in ClickUp
SEMICOLON - SE - 62 | ix
Declaration
Figure 31 : Collaboration between G-Drive and Slack

Figure 32 : Snip of allocations for the weekly meeting schedule
Figure 33 : Snips of Survey
SEMICOLON - SE - 62 |x
Declaration
LIST OF TABLES
Table 1 : Comparison between existing products

Table 2 : Research Objectives
Table 3 : Academic Objectives
Table 4 : Operational Objectives
Table 5 : Hardware Requirements
Table 6 : Required Programming Language Software
Table 7 : Required IDE’s and Other Software
Table 8 : Required APIs and Frameworks
Table 9 : Technology Stack
Table 10 : Research philosophies comparison
Table 11 : Comparison of the research methodologies
Table 12: Comparison of the development methodologies
Table 13 : Comparison of design methodologies
Table 14 : Comparison of evaluation methodologies
Table 15 : Risks and Mitigation
Table 16: The Stakeholder descriptions table
Table 17:Advantages,Disadvantages and findings of the literature review
Table 18 :Advantages,Disadvantages and findings of the literature review
Table 19 :Advantages,Disadvantages and findings of the interviews
Table 20 :Advantages,Disadvantages and findings of brainstorming
Table 21 :Advantages,Disadvantages and findings of Prototyping
Table 22: Questionnaire Findings table
Table 23 :Interview 1
Table 26: Selection/ findings of Literature Review
Table 27 :Use Case descriptions for “Authentication”
Table 28 :Use Case descriptions for “Ask Questions”
Table 29 :Use Case descriptions for “Visualize”
Table 30 :Use Case descriptions for “Breakdown”
SEMICOLON - SE – 62 | xi
Declaration
Table 31 :Use Case descriptions for “Rephrasing”

Table 32 :Use case descriptions for “Post questions”
Table 33 :Use Case descriptions for viewing IEP report
Table 34 :The Functional Requirements table.
Table 35 :The Non Functional Requirements table
Table 36 : Descriptions of classes in Class Diagrams
Table 37 : Schedule for weekly meetings and additional sessions
Table 38 : Work Breakdown Matix table
SEMICOLON - SE – 62 | xii
Abbreviations Table
LIST OF ABBREVIATIONS
Abbreviation Explanation
AI Artificial Intelligence
APA American Psychiatric Association
API Application Programming Interface
ASD Autism Spectrum Disorder
BCS British Computer Society
BERT Bidirectional Encoder Representations from Transformers
CDC Centers for Disease Control and Prevention
DL Deep Learning
DSM Diagnostic and Statistical Manual of Mental Disorders
FAST Families and Schools Together
GE General Education
GLIDE Guided Language to Image Diffusion for Generation and Editing
GPT Generative Pre-trained Transformer
HTTP Hyper Text Transfer Protocol
IEP Individualised Educational Plans
JS JavaScript
KPMG Klynveld Peat Marwick Goerdeler
LR Literature Review
ML Machine Learning
NASS National Autism Spectrum Disorder Surveillance System,
NGO Non-governmental Organisations
NLP Natural Language Processing
NLTK Natural Language Toolkit
OOADM Object Oriented Analysis Design Method
SEMICOLON - SE - 62 | xiii
Abbreviations Table
PHA Public Health Agency
RNN Recurrent Neural Network
RoBERT Robustly Optimized BERT
SEN Special Educational Needs
SLEP Social, Legal, Ethical and Professional Issues
SMS Short Message Service
SRS System Requirements Specification
SSADM Structured System Analysis Design Method
UI User Interface
UNCRPD United Nations Convention on the Rights of Persons with Disabilities
UNESCO United Nations Educational, Scientific and Cultural Organization
UX User Experience
VR Virtual Reality
WBS Work Breakdown Structure
SEMICOLON - SE - 62 | xiv
CHAPTER 1 - Introduction
CHAPTER 1: INTRODUCTION
1.1 Chapter Overview
The introduction chapter provides a comprehensive overview of the project being undertaken. To
begin with, the project background is concisely described, offering context for the work to follow.
This is followed by a thorough exploration of the problem background, delving into the motivations
and challenges involved. The significance of the project is then addressed, highlighting the potential
impact and relevance of the work. Previous research in the field is carefully examined, offering a
foundation for the project and identifying any gaps that the project aims to fill.
The objectives of the project are clearly defined, serving as a roadmap for the work to come. These
objectives outline the specific goals that must be achieved in order to successfully complete the
project. The scope of the project is also defined, outlining the boundaries of the work and specifying
what is and is not included.
Finally, the resource requirements for the project are discussed in detail, including the hardware,
software, and data resources that will be necessary. This ensures that the project has the necessary
support and resources to be successful.
1.2 Problem Background
1.2.1 Autism And Inclusive Education
Autism1 is a neurodevelopmental disorder that is characterised by a range of behavioural, social, and

communication difficulties that can vary in severity (American Psychiatric Association [APA], 2013).
These difficulties may manifest in the form of restricted or repetitive behaviors, interests, or activities,
and can impact an individual's functioning in multiple contexts. The prevalence of autism is estimated
to be around 1 in 93 children in Sri Lanka (Hossain et al., 2017). The rate of diagnosis has been on
the rise, for example in Canada as many as 1 in 66 children receiving a diagnosis (Hoffman, 2013;
National Autism Spectrum Disorder Surveillance System, 2018). The National Autism Spectrum
1
Throughout this study, we will use the words "autistic" and "autism" interchangeably, as recommended by self-advocates
and in line with current views on language and labeling (Canadian Autism Spectrum Disorder Alliance, 2020; Kapp et
al., 2013; Nicolaidis et al., 2016; Robinson, 2019).
SEMICOLON - SE - 62 |1
Disorder Surveillance System [NASS] (2018) reports that the prevalence of autism in Canada has
increased over the past decade.
Despite the increasing prevalence of autism, there has been relatively little research on students with
autism in inclusive classrooms (McKeating, 2013). As more children and youth with autism are
educated in inclusive classrooms (WHO, 2017; Geneva Centre for Autism, 2010), general education
teachers are responsible for supporting their success in these environments (Hoffman, 2013).
However, despite the adoption of inclusive education policies in many countries around the world,
including Sri Lanka, the implementation of inclusive education in practice has been slow and
inconsistent, with many students with disabilities continuing to be educated in separate special
education settings or excluded from educational opportunities altogether (Hettiaarachi et al., 2017).
In Sri Lanka, the slow implementation of inclusive education and the associated stigma, including
parental disbelief of diagnoses, were cited by teachers as barriers to its effective implementation (SEN
Teacher, 2022; Trainee SEN Teacher, 2022). These challenges make it difficult for educators to take
the necessary steps to adapt lesson plans for the benefit of all students.
Autistic students may experience a range of cognitive, learning, speech/language, and communication
difficulties in addition to their autism diagnosis (Mashal & Kasirer, 2011; Semrud-Clikeman et al.,
2014). These students may also exhibit problematic behaviors such as tantrums, aggression, or
destructiveness, which can arise from a variety of factors including difficulties with communication
and adapting to change (Ontario Ministry of Education, 2007). In order to support the development
of students with autism in the educational environment, they may require specialized supports and
accommodations (Mandell et al., 2005; Mashal & Kasirer, 2011; Semrud-Clikeman et al., 2014).
1.2.2 Teaching in Inclusive Education
Teachers are responsible for implementing strategies to manage the cognitive, academic, behavioral,
and social difficulties of students with autism, which can significantly increase the complexity and
workload of these teachers (Lindsay et al., 2014; Marks et al., 2003). The success of students with
autism in the classroom is largely dependent on the strategies implemented by teachers (Hart &
Whalon, 2011). However, there are several challenges to the implementation of inclusive education
in Sri Lanka that can make it difficult for teachers to effectively support the success of these students
in inclusive classrooms. Many teachers in Sri Lanka reported a lack of access to training and
professional development opportunities related to inclusive education, which can make it difficult for
them to effectively support the inclusion of students with disabilities in mainstream classrooms.
Educators within the field have corroborated that a scarcity of resources has been a persistent problem,
leading to feelings of frustration due to their inability to provide optimal teaching support for their
students. Additionally, the need to constantly employ alternative methods has been perceived as
inconvenient and time-consuming, particularly in comparison to the more efficient teaching strategies
utilized in other countries (SEN Teacher, 2022; Trainee SEN Teacher, 2022).
Another challenge is the prevalence of stigma and discrimination against individuals with disabilities
in Sri Lanka, including negative attitudes and beliefs held by teachers, administrators, and members
of the community, as well as discrimination in the form of segregation or exclusion from educational
opportunities (Hettiaarachi et al., 2017; Sciaraffa et al., 2019). Sciaraffa et al. (2019) found that
disability stigma can have a significant impact on the lives of individuals with disabilities, including
reduced access to education, employment, and other opportunities.
There is also a lack of awareness about the benefits of inclusive education among parents,
administrators, and members of the community in Sri Lanka (Hettiaarachi et al., 2017), which can
lead to resistance to inclusion and a lack of support for teachers who are attempting to implement
inclusive practices. These challenges can make it difficult for GE teachers to effectively support the
success of students with autism in inclusive classrooms.
1.2.3 Technologies to Support Teaching in Inclusive Education
The use of technology, particularly software applications, in teaching children with autism can be an
important tool in supporting their learning and development. This is due to several factors. Firstly,
children with autism often have unique learning needs that may be better supported by technology,
such as visual or auditory aids (Anoyiannakis, 2013) or opportunities for individualized instruction
(Ghanouni et al., 2019). Secondly, technology can be an effective way of engaging children with
autism in the learning process, such as through interactive learning (Anoyiannakis, 2013). Finally,
technology can be a valuable tool in helping children with autism to develop important skills such as
communication and problem-solving (Ghanouni et al., 2019). However, there are several barriers to
the adoption of technology in the classroom, including a lack of funding and resources (Ghanouni et
al., 2019), a lack of research on the effectiveness of these technologies (Anoyiannakis, 2013), and a
lack of awareness or understanding among teachers (Gómez-Marí et al., 2022). These issues may
prevent teachers from utilizing technologies to teach children with autism.
The following table illustrates significant applications that have been found to aid teachers in
effectively managing their time in the classroom and in gaining a deeper understanding and
readiness for an improved teaching experience for their students'
Product/ Feature Frontline Cough Avanz LiftEd SupAut

Education Drop
Visualizer
Breakdown instructions
Teach social conversing standards
Prepare IEP report
Automate IEP reports
Organise child progress in a

dashboard
Organise assignments and answer

reviews systematically
Conversation assignment helper
Table 1 : Comparison between existing products
Project SupAut is a unique technology that combines the best features of the other technologies to
effectively bridge the gap between management and application. The management system helps
teachers track student progress and efficiently utilize their time in the classroom in applications such
as Frontline Education and LiftEd. While in applications such as CoughDrop and Avanz the use of
communication and assistive technology which enables teachers to gain valuable insights into a
child's development and identify areas where they may need additional support, is a source of
inspiration to the making of SupAut. By combining these two applications and adding additional
assessments and questions, Project SupAut offers a comprehensive solution for teachers working with
children with autism, improving both academic and other skills such as social, logical, creative, and
emotional skills.
1.3 Problem Statement
Autistic students in traditional classroom settings often face challenges in social, creative, and
communication skills and the documentation of their progress can be time-consuming and resource-
intensive. There is a need to identify more efficient and effective methods and tools for tracking and
documenting the progress of these students and to explore the potential of assistive technology to
support language and communication development.
1.4 Research Gap
The research gap in this area involves identifying effective methods and tools for tracking the social,
creative, and communication skills of autistic students in traditional classroom settings. This requires
the development of standardized and objective methods for measuring and assessing these skills, as
well as strategies for collecting and analyzing data over time.
There is also a gap in determining the most efficient method for automating the documentation of
progress in academics and other skills for autistic children, which involves evaluating various
technology solutions or approaches in terms of factors such as cost, ease of use, and reliability. This
is very important as many educators reportedly feel overwhelmed in the preparation of IEP reports
(Carr et al., 2002).
Additionally, there is a need to understand how assistive technology tools can be designed and
implemented to effectively support autistic children with language and communication challenges in
completing homework assignments independently, especially since it has been proved that it helps
autistic childre maintain attention span while motivate them in the learning process (Gómez-Marí et
al., 2022). Anoyiannakis (2013) also stresses on the vital role technology plays in providing
opportunities for interactive learning, which can be more engaging for children with autism than
traditional, lecture-based instruction. This involves researching the needs and challenges faced by
these children, as well as exploring different types of assistive technology and their potential for
supporting language and communication skills.
1.5 Research Questions
The main goal of this research project is to address the following research arguments:
RQ1: What methods and tools can teachers use to systematically track the social, creative, and
communication skills of their autistic students in a traditional classroom setting?
RQ2: What is the most efficient method for automating the documentation of progress in academics
and other skills for autistic children?
RQ3: How can assistive technology tools be designed and implemented to help 7-12 year-old autistic
children with language and communication challenges complete homework assignments
independently?
1.6 Research Aim
To design and develop an effective method for tracking and documenting the progress of autistic
students in traditional classroom settings, with a focus on social, creative, and communication skills,
to test and evaluate the potential of assistive technology to support language and communication
development in 7-12 year-old autistic children.
This research aim has two main components:
● Identifying effective methods and tools for tracking and documenting the progress of autistic
students: The research aims to identify methods and tools that teachers can use to
systematically track the progress of their autistic students in terms of social, creative, and
communication skills in a traditional classroom setting. This may involve researching existing
approaches and tools, as well as developing and testing new ones.
● Exploring the potential of assistive technology to support language and communication

development in 7-12 year-old autistic children: The research also aims to explore the potential
of assistive technology to help 7-12 year-old autistic children with language and
communication challenges complete homework assignments independently. This may
involve identifying and evaluating existing assistive technology tools, as well as designing
and testing new ones.
Overall, the research aim is focused on finding ways to improve the tracking and documentation of
progress for autistic students and to support their language and communication development through
the use of assistive technology. It is specifically targeting traditional classroom settings and 7-12 year-
old autistic children, and is focused on social, creative, and communication skills.
1.7 Project Scope
The scope of the project is focused on designing and implementing assistive technology tools that
enable 7-12 year-old autistic children with language and communication challenges to complete
homework assignments independently. This will ultimately assist teachers in systematically
monitoring the social, creative, and communication skills of their autistic students in a traditional
classroom setting, with a focus on core functionalities and techniques in the initial stage. The scope
of the project is based on project objectives and a review of the relevant subject area and existing
products.
1.7.1 In-scope
The prototype currently includes the following features at the start of the project:
● Ability to interact and view analytics of multiple student in a class, through cross examination
● Conversing with students to both, send assigned questions as well as answer any doubts raised
by the student, through the process of submitting the assignment, using conversational AI
tools
● Life-like interaction to replicate a teacher in the classroom, with the ability to break down
complex questions in parts through which the child can better understand, with the help of
transformer-based language models.
● Assists in visualising concepts and entities through the use of AI generative art.
● Detect and correct possible anti-social conversing standards, through the use of sentiment
analysis of the Natural language processing (NLP) technique
1.7.2 Out-scope
● At the initial attempt, it will be challenging to train the chatbot to converse in multiple
languages. However, we plan to offer multi-language support as a future enhancement.
● Initially, we will be integrating the chatbot to work exclusively with the WhatsApp messenger
application.
● The dashboard has been designed with the assumption that a teacher is only teaching one
subject at a time.
1.8 Rich Picture Diagram
Figure 1 : Rich Picture Diagram of Project SupAut (Self - Composed)
1.9 Objectives
1.9.1 Research Objectives
RO1 Research a suitable way to Identify children affected with Autism
Description Aim to learn existing techniques and solutions of identifying the

children who are affected with autism
RO2 Research a suitable way of guiding the Autistic Child
Description Aim to conduct deep research to sort out what are the most
effective techniques which could help in the guidance
RO3 Research about how can the technology be included in the

Identification Process of Autism
Description Aim to conduct in-depth research in order to find the ways to fit
technology in to the Identification process
RO4 Research about how can the existing technologies be included in

order to guide the Autistic child
Description Aim to study the existing and new technologies in guiding an

Autistic child
RO5 Research Limitations and future improvements of the project
Description Being aware of the limitations of this specific research concerning

to focus on the objectives . Therefore the findings can be stored
for future researches
Table 2 : Research Objectives
1.9.2 Academic Objectives
AO1 Learn how a project development works
Description Developing the project calls for teamwork skills. This project
offers the right set of circumstances to research and understand
those requirements; experience developing, documenting and
implementing real world solutions.
AO2 Learn to design a software system model
Description Designing a solid software model through applying theoretical

understanding of software design to create a model solution is the
academic objective.
AO3 Learn to use Machine Learning technologies
Description This project uses Machine Learning techniques which are relevant
to the Modules upcoming. Therefore gaining knowledge about
this is an academic objective
AO4 Learn to use relevant resources properly.
Description Learning to use the available relevant resources properly and

effectively is an academic objective
AO5 Learn to develop the software
Description Overall developing a completely functional software program

holds almost all of the academic objectives. It offers the
possibility to look at the entire view of the operations undertaken
withinside the software both backend and frontend development,
implementation, organisation of code and more.
Table 3 : Academic Objectives
1.9.3 Operational Objectives
OO1 Identify and analyse research components
Description Do a research in Autism Spectrum Disorder (ASD or Autism) to

identify the components and analyse it deeper to develop research
components
OO2 Pin down the requirements and collect data
Description Use different techniques to determine the must include

requirements to develop the system. Collect data and validate all
doubts and receive input from system related individuals (domain
experts , mentors, etc )
OO3 Find and analyse existing solutions
Description Conduct a research on the existing solutions which are related ,

critically analyse , understand the techniques and identify it’s
pros and cons
OO4 Design the Project
Description Identify all the necessary Components which are required in the
designing phase. Critically understand the scenarios with the help
of case studies , choose the most suitable software and approach
to design the system
OO5 Implement the Project
Description Using suitable programming languages, algorithms , APIs , etc in

the phase of implementing the complete system
OO6 Test and evaluate the System
Description Testing the system to fix errors , bugs and update the system .
Checking the Final System and ensuring if the system fulfils all its
requirements.
SEMICOLON - SE - 62 | 10
Table 4 : Operational Objectives
1.10 Resource Requirements
1.10.1 Hardware Requirements
Hardware
Intel Dual Core or above This type of processors are required to work
Ryzen 3 or above with the software components
8GB 1600MHz RAM or above The software runs various components

simultaneously. Therefore the mentioned
RAM is needed at least
More than 20GB Disk Space Managing reports , Analysing data ,

installation of required software and other
Data related processes will need more
storage space
GPU higher than 2GB Working with ML and DL models need this
Table 5 : Hardware Requirements
1.10.2 Software requirements
Programming Languages
Node JS, For the ML Programming
HTML5 , CSS , JavaScript , React, For the web Development

CoreUI Library
React JS For the creation of hybrid software

(Responsive for both web and mobile
platforms )
Table 6: Required Programming Language Software
IDEs and Other Software
Visual Studio Code For front-end development code editor
Node JS For back-end development
Firebase For back-end development and Hosting
Figma For the designing of prototypes and

wireframes
Google Docs For Documentation and formatting of

reports
Draw.io , Lucidchart For drawing the diagrams
Google Drive For file management , documenting and

storage
Git For version control
Slack For communication with the Teammates
ClickUp For Project and time Management
Table 7 : Required IDE’s and Other Software
APIs and Frameworks
GPT-3 API For optimising language model for dialogue
Twilio REST API Query and manage meta-data of the project
Twilio (Messaging , Whatsapp, For the Chatbot Integration

Programmable chat APIs )
Twilio Sync API For Syncing data and build live-updating

Dashboards
Dall - E AI art generation
Table 8 : Required APIs and Frameworks
1.10.3 Technology Stack
Front-End
Back-End
NLP Platform
Google Cloud Natural Language Open AI GPT-3 Open AI DALL-E
Infrastructure
Version Control
Services
Table 9 : Technology Stack
1.11 Chapter Summary
This chapter presents a summary of the methods and technologies that can be employed to track the
progress of autistic students in traditional classroom settings. These techniques aim to address the
difficulties that these students often experience with social, creative, and communication skills, and
to make the process of documenting their progress more efficient. The chapter also provides a
thorough examination of the problem domain and offers background information on the project. The
project's goals and objectives are outlined, as is the scope of the project and the necessary resources
required for its implementation. The following chapter examines previous research and approaches
in the field, as well as the selected approaches for the SupAut project.
CHAPTER 2 - Literature Review
CHAPTER 2: LITERATURE REVIEW
In the previous chapter, the project background, the problem domain was overviewed, and how
Project SupAut intended to provide a solution. However, this chapter will delve deeper into the
problem domain and critically evaluate the previous research conducted by other researchers to
understand the approaches, techniques, and technologies that are currently available and their
advantages and disadvantages. The goal of this analysis is to determine the best approach for
implementing Project SupAut. The final selected approach will also be explained in this chapter,
along with an in-depth review of the techniques used for evaluation.
2.2 Concept Graph
In the first phase of the literature review, a concept graph was created to plan and organize the
chapter's content. The areas of focus were determined based on the domain of the project, technology
reviews, and reviews of existing systems. This approach was used to systematically gather literature
and ensure that all necessary areas were covered. The concept graph can be found in the appendix,
Section A.1.
2.3 Domain
2.3.1 Autism
Over the years, the understanding and treatment of autism have evolved significantly, as reflected in
successive editions of the Diagnostic and Statistical Manual of Mental Disorders (DSM) (American
Psychiatric Association [APA], 1980; 1994; 2000; 2013). According to the current DSM-5 definition,
it is a condition that affects brain development, resulting in difficulties with social interaction and
communication, as well as repetitive behaviors that can impact an individual's daily functioning. It is
classified as a neurodevelopmental disorder. (APA, 2013). The mental abilities and social
communication skills of individuals with autism may vary. (APA, 2013), and their symptoms may
change over time (APA, 2013).
2.3.1.1 Prevalence
According to recent studies, autism is a prevalent neurodevelopmental disorder that impacts around
1 in 59 children in the United States. (Dowling et al., 2020; Centers for Disease Control and
Prevention [CDC], 2018). In Sri Lanka, the prevalence of autism is estimated to be around 1 in 93
children (Hossain et al., 2017). Boys are diagnosed with autism at a higher rate than girls, diagnosed
five or four times as frequently as usual (Schaefer, 2016). The prevalence of autism has significantly
increased in recent years, with previous estimates putting the number at 1 in 150 children (CDC,
2018). This increase can be attributed to several factors, including increased public awareness,
availability of resources, and improved diagnostic practices (King & Bearman, 2009; VanBergeijk et
al., 2008).
2.3.1.2 Diagnostic Criteria
According to the Diagnostic and Statistical Manual of Mental Disorders (DSM; American Psychiatric
Association [APA], 2013). People with autism often have difficulty with social communication and
engage in repetitive behaviors that can affect their daily lives. These impairments and behaviors must
be present for an individual to be diagnosed with autism. These challenges may be present from an
early age (Hyman et al., 2020).
People with autism may struggle with social communication, including engaging in back-and-forth
dialogue, initiating social interactions, or responding emotionally to others. This may involve
difficulties with mutual emotional expression and understanding or a lack of social-emotional
reciprocity. They may also have atypical nonverbal communication skills, such as difficulty with
body language, eye contact, and gestures, or a lack of facial expressions. Additionally, individuals
with autism may struggle with different social interactions making it challenging to develop,
maintain, and comprehend peer relationships, for example making friends, sharing with and showing
interest in others(APA, 2013), this problematic nature of social communication is an area of curiosity
that Project SupAut seeks to address through diligent effort.
Some individuals with autism may exhibit an insistence on maintaining consistency and routine, have
very specific and focused interests, and have a heightened sensitivity to sensory stimuli. These
behaviors may involve repeating the words of others, using objects in a specific way, having inflexible
routines, having intense interests in particular objects or topics, or reacting unusually to sensory
stimuli (APA, 2013). To receive a diagnosis of autism, individuals must exhibit at least two of these
behaviors in the restrictive and repetitive behaviors domain. It is important to keep in mind that the
DSM criteria for diagnosing autism are constantly being updated and refined, and it is essential to
consult with a qualified mental health professional to accurately diagnose and understand the specific
needs and challenges of an individual with autism (Pagaling & Rachel, 2020).
2.3.2 The Inclusive Education
It is a type of education that values diversity and promotes the participation and belonging of all
students, including those with disabilities or diverse backgrounds. It involves the provision of
accommodations and modifications as needed, and collaboration between teachers, students, and
families to support the learning and development of all students. The foundation of inclusive
education is equity and respect for the rights and dignity of all students (UNESCO, 1970).
2.3.2.1 Integration and Mainstreaming
Compared to their neurotypical peers, students with autism often encounter more difficulties in the
academic setting. This can impact their educational and social experiences (Brue et al., 2018). Autistic
students may be more likely to be bullied by their peers, have difficulty socializing, and struggle to
work collaboratively (Brede et al., 2017; Kurth & Mastergeorge, 2012). It is important to note that
autistic individuals may experience a wide range of challenges in school and may require additional
support to succeed academically and socially, which could lead to a higher possibility of other mental
illnesses like anxiety and depression (DeFilippis, 2018; Strang et al., 2012). These challenges may be
exacerbated by sensory differences that make it harder for autistic students to cope in school settings
(Brue et al., 2018). In addition to that it is recorded that, autistic students are at an increased risk of
being expelled from school compared to their non-autistic peers (Brede et al., 2017).
Earlier, it was mentioned that research has demonstrated that students with autism frequently
need additional learning and social support compared to their peers without autism (Mandell
et al., 2005), and those with more severe symptoms may need even more assistance (Nuske
et al., 2019). With the increasing prevalence of autism, it is expected that more children with
autism will be placed in inclusive classrooms (Canada, PHA, 2018). It is important to
consider the unique needs and challenges faced by these students and to provide them with
the necessary support to succeed in their academic and social environments.
2.3.2.2 Inclusive Education in Sri Lanka
Inclusive education is a model of education that aims to provide an equal educational

experience to all students, including those with disabilities. Inclusive education has gained
widespread recognition as a best practice in education. The UNCRPD, which was adopted
by the United Nations in 2006, affirms the right of all individuals with disabilities to receive
inclusive education. Despite the adoption of inclusive education policies in many countries
around the world, including Sri Lanka, the implementation of inclusive education in practice
has been slow and inconsistent, with many students with disabilities continuing to be
educated in separate special education settings or excluded from educational opportunities
altogether (Hettiaarachi et al., 2017). The slow implementation of inclusive education in Sri
Lanka and the associated stigma, including parental disbelief of diagnoses, were repeatedly
cited by teachers interviewed for this paper as hindrances to its effective implementation.
These challenges make it difficult for educators to take necessary steps to adapt lesson plans
for the benefit of all students (SEN Teacher, 2022; Trainee SEN Teacher, 2022).
2.3.2.2.1 Challenges to the Implementation of Inclusive Education in Sri Lanka
One key challenge to the implementation of inclusive education in Sri Lanka is the lack of
resources and support for teachers who are responsible for implementing this model of
education. This includes a lack of trained personnel, specialized equipment, and funding to
support the additional accommodations or support that students with disabilities may need.
For example, many teachers in Sri Lanka reported a lack of access to training and professional
development opportunities related to inclusive education, which can make it difficult for them
to effectively support the inclusion of students with disabilities in mainstream classrooms. It
has been corroborated by educators within the field that a scarcity of resources has been a
persistent problem, leading to feelings of frustration due to their inability to provide optimal
teaching support for their students. Additionally, the need to constantly employ alternative
methods has been perceived as inconvenient and time-consuming, particularly in comparison
to the more efficient teaching strategies utilized in other countries (SEN Teacher, 2022;
Trainee SEN Teacher, 2022).
In Sri Lanka, one of the main obstacles to implementing inclusive education is the widespread
stigma and discrimination faced by people with disabilities. This may include negative
attitudes and beliefs held by teachers, administrators, and members of the community, as well
as discrimination in the form of segregation or exclusion from educational opportunities
(Hettiaarachi et al., 2017; Sciaraffa et al., 2019). One study, conducted by Sciaraffa et al. in
2019, discovered that disability stigma can significantly affect the daily lives of individuals
with disabilities, including reduced access to education, employment, and other opportunities.
In addition to these challenges, teachers in Sri Lanka may also face difficulties in
implementing inclusive education due to a lack of awareness about the benefits of this model
of education among parents, administrators, and members of the community (Das et al.,
2018). This may lead to resistance to inclusion and a lack of support for teachers who are
attempting to implement inclusive practices.
2.3.2.3 Approaches to Inclusive Education
It is a diverse and multifaceted concept that can be interpreted and implemented in different
ways depending on the context (Alquraini & Rotatori, 2014; Slee, 1998), while the reality of
its implementation can vary significantly (Alquraini & Rotatori, 2014). For example, in
Alberta, the government has mandated that regular classrooms should be the first option for
placement for all students with special needs and that teachers should adapt their instruction
to ensure that all students, including those with exceptional learning needs, can achieve success
(Alberta Education, 2017; McCrimmon, 2015; McLeskey et al., 2012). Additionally, students
with more severe needs are often placed in separate settings, such as separate classrooms
(Graham & Sweller, 2011; Pulkkinen & Jahnukainen, 2015). This may be due in part to a lack
of clear guidelines and training for educators on inclusive education practices, leaving them
with the challenge of making inclusion work in their classrooms (Dogan et al., McCrimmon,
2015). The lack of support from parents, including instances of refusal to accept a diagnosis,
has been identified as a significant challenge that exacerbates the difficulty of the situation
(SEN Teacher, 2022; Trainee SEN Teacher, 2022).
2.3.2.4 Inclusion for Students with Autism
Inclusive education is a diverse and multifaceted concept that can be interpreted and
implemented in different ways depending on the context ( Alquraini & Rotatori, 2014; Slee,
1998). In Alberta, the government has mandated that regular classrooms should be the first
option for placement for all students with special needs and that teachers should adapt their
instruction to ensure that all students, including those with exceptional learning needs, can
achieve success in meaningful ways (Alberta Education, 2017; McCrimmon, 2015;
McLeskey et al., 2012). However, research suggests that full inclusion has yet to be fully
realized in many cases, and students with special needs may still be separated from their
peers and following their curriculum, with limited social or academic interactions (McLeskey
et al., 2010; Sailor, 2014).
The inclusion of students with autism in mainstream classrooms has been found to have
positive effects on their social interactions, levels of support, and educational goals compared
to those in segregated settings (Eldar et al., 2010). Inclusive classrooms can provide a
supportive and stimulating environment for all students to learn and grow alongside their peers
(Chandler-Olcott & Kluth, 2009; Zhang & Hu, 2015). Project SupAut aims to automate the
process of providing additional support for children with the recognition that inclusive
education can present challenges to teachers.
2.3.3 Teaching in Inclusive Education
2.3.3.1 Challenges in teaching in an Inclusive Classroom
Teaching students with autism in an inclusive classroom setting can be difficult for teachers.
This is because autistic students often have unique needs and may require special
accommodations and support to fully participate in the classroom (Sansosti et al., 2016).
General education teachers often been reported feeling unprepared and overwhelmed by the
demands of including students with autism in their classrooms. These feelings may be
intensified by a lack of resources or support from specialists, as well as a lack of understanding
or acceptance from colleagues and administrators (Al-Sharbati et al., 2013). Inclusive
education can present significant challenges for teachers, especially when it comes to meeting
the varied needs of each child in the classroom, which is the underlying preface for the project.
Exhaustion, stress, and burnout among general education teachers have been correlated with
their understanding of effective methods for managing difficult student behaviors and lack of
preparation (Coman et al., 2013; Gray et al., 2017). Both the SEN Teacher (2022) and the
Trainee SEN Teacher (2022) interviewed, mentioned this multiple times in their interviews.
The extra effort required of teachers, such as after-class preparation (e.g. IEP reports) and
tracking the progress of individual students, can lead to feelings of being overwhelmed and
exhausted.
Teachers have often reported facing several difficulties when teaching students with autism,
including a lack of confidence in their ability to teach these students, the time-consuming
nature of collaborating with other educators and school staff, concerns about managing
challenging behaviors in the classroom, difficulties with individualized education plans
(IEPs) and data collection (Lindsay et al., 2014). These challenges formed the basis for the
SupAut project, which sought to identify the most effective solutions to these issues and
evaluate the prevalence and significance of these problems.
Furthermore, preparing IEP reports can be a challenging and time-consuming process for
teachers, particularly for those who are not familiar with the process or who do not have
access to resources or support (Carr et al., 2002). A study by Carr et al. (2002) found that
teachers often report feeling overwhelmed by the demands of preparing IEP reports,
particularly when they are required to gather and organize a large amount of information
about a student's needs and progress. This can be particularly challenging for teachers of
students with autism, as these students may have complex needs and may require multiple
accommodations and support to participate fully in the classroom (Carr et al., 2002)
Teachers in inclusive classrooms may face several challenges when it comes to evaluating
extra homework and not being able to administer the work of the child from home. These
challenges may include a lack of time or resources to provide additional support or feedback
on homework assignments (Merriman et al., 2016), difficulty in accurately assessing the
progress of students with disabilities (Kameenui, 1998), and limited opportunities for
collaboration or communication with parents or caregivers (Families and Schools Together
[FAST], 2002).
A study by Merriman et al. (2016) found that teachers often report feeling overwhelmed by
the demands of grading and providing feedback on homework, particularly when they have a
large number of students or when the assignments are complex or time-consuming. This can
be especially challenging for teachers of students with disabilities, as these students may
require additional support or accommodations to complete their homework (Merriman et al.,
2016). In addition, teachers may find it difficult to accurately assess the progress of students
with disabilities, particularly if they do not have access to appropriate assessment tools or
resources (Kameenui, 1998).
2.3.3.2 School Resources in an Inclusive Classroom
Teachers often report a lack of resources such as staff development training, preparation
programs, administrative support, collaboration, co-teaching, support, and funding (Busby et
al., 2011; Ruel et al., 2015). These resources are important for teachers to have a good
understanding of autism to effectively teach and support students with autism in the classroom
(Sansosti et al., 2016; Conderman & Johnson-Rodriguez, 2009; Ross Hill, 2009). In
particular, local teachers have been noted as being behind in the use of efficient methods used
in other parts of the world (SEN Teacher, 2022, Rifkhan, 2022). This can be frustrating for
teachers who are aware of better and more effective practices that could be implemented but
are unable to access these resources due to a lack of funding and support.
The involvement of other staff members, such as administrative assistants, classroom aides,
and special education teachers, in the education of students with autism can be a valuable
resource for general education (GE) teachers (Guldberg, 2010; Lindsay, 2013; Symes &
Humphrey, 2011). Ainscow (2007) recommends promoting a collaborative culture within
schools, where teachers and support staff work together and communicate to create a more
inclusive environment. Collaboration between special education teachers and GE teachers
has been found to enhance the understanding and capacity of GE teachers in teaching students
with autism (Da Fonte & Barton-Arwood, 2017; Pancsofar & Petroff, 2013). This
collaborative approach has been found to boost the confidence of teachers and encourage
them to seek further collaboration with school staff (Pancsofar & Petroff, 2013). However,
this aspect of collaboration among teachers can also be time-consuming and exhausting,
requiring coordination to ensure all teachers are up-to-date on the student's progress. Easing
collaboration and co-teaching through improved communication and coordination can also
be an important part of this paper.
2.4 Technologies
2.4.1 Selection of AI Art Generator
It has been found through research that individuals with autism tend to process and retain
information better when it is presented visually. Therefore, inclusive education teachers often
find it helpful to include visual elements in their lesson plans (Erpelding, 2013; Tissot &
Evans, 2003). While this strategy can be effective, it can also be time-consuming and require
a lot of preparation. Project SupAut aims to address this issue by utilizing artificial
intelligence (AI) to generate visual aids more efficiently.
2.4.1.1 Choosing the Right Conversational AI for Project SupAut
DALL-E is a state-of-the-art AI art generator that has gained widespread recognition for its
ability to generate unique and creative artworks (OpenAI, 2021). In comparison to other AI
art generators such as Midjourney and Imagen, DALL-E has several advantages that make it
a better selection.
Firstly, DALL-E uses a novel approach to generating art, using a combination of

unsupervised and supervised learning techniques to generate original and diverse artwork
(OpenAI, 2021). In contrast, Midjourney and Imagen primarily rely on supervised learning
techniques, which can limit the range and creativity of their output (Zhang, L. et al., 2020).
Secondly, DALL-E has a large and diverse dataset, comprising over 50 million images and
texts, which allows it to generate artworks that are more realistic and diverse (OpenAI, 2021).
In comparison, Midjourney and Imagen have smaller datasets, which may result in less
realistic and diverse artworks (Zhang, L. et al., 2020).
Finally, DALL-E has been extensively tested and validated by researchers, with several
studies demonstrating its ability to generate high-quality and creative artwork (OpenAI,
2021). In contrast, there is limited research available on the performance of Midjourney and
Imagen, making it difficult to assess their effectiveness as AI art generators (Zhang, L. et al.,
2020).
Overall, the research supports the selection of DALL-E as the superior AI art generator in
comparison to Midjourney and Imagen. DALL-E's unique approach to generating art, large
and diverse datasets, and extensive validation make it a more reliable and effective choice for
generating original and creative artwork.
2.4.1.2 DALL-E- Under the Hood
According to the research paper "DALL-E 2: Generating High-Quality Images from Text
Descriptions" (Radford et al., 2021), the DALL-E 2 image-generation process involves the
following steps:
Figure 2 : DALL-E 2 image-generation process (Ramesh, A. et al, 2022)
● Receiving input: DALL-E 2 receives a text description of an image as input (Radford

et al., 2021).
● Text-to-image generation: DALL-E 2 uses a transformer-based architecture (Radford
et al., 2021) to generate an initial image based on the input text description.
● Image evaluation: The initial image is then evaluated using a scoring function
(Radford et al., 2021) that measures its quality, including factors such as sharpness,
diversity, and alignment with the input text description.
● Iterative improvement: If the initial image does not meet a certain quality threshold,
DALL-E 2 uses a feedback loop (Radford et al., 2021) to iteratively improve the
image by adjusting the image's pixels and re-evaluating the image until it meets the
quality threshold.
In addition to these steps, the DALL-E 2 image-generation process also involves the use of
two techniques called "clip training" and "GLIDE" (Radford et al., 2021). Clip training
involves using a loss function that only backpropagates through the pixels of the image that
are above a certain threshold, which helps to prevent the model from generating blurry images
(Radford et al., 2021). GLIDE is a method for resizing the images during training, which
helps to reduce the amount of computation required and improve the model's performance
(Radford et al., 2021).
Figure 3: A diagram that illustrates the Diffusion Model (Ramesh, A. et al, 2022)
The DALL-E 2 image-generation process also utilizes the diffusion model (Rogers, 2003),
which is a way of describing how new ideas or innovations spread through a population. The
use of the diffusion model in DALL-E 2 allows the program to learn from and mimic the way
that humans adopt and spread new ideas (Radford et al., 2021). By using the diffusion model,
DALL-E 2 can better understand how humans respond to and adopt new concepts and ideas,
which can help it to generate more human-like responses and generate more creative ideas
(Radford et al., 2021).
2.4.2 Selection of Conversational AI
It is important for building an efficient chatbot because it allows the chatbot to understand and
respond to user input in a natural, human-like manner. This is important because users expect
a high level of responsiveness and personalization from chatbots (Koufaris & Hamari, 2018).
Chatbots that use artificial intelligence are also rated as more helpful and efficient than those
that do not (KPMG, 2021). The efficiency, effectiveness, and responsiveness of the chatbot
are of main functional importance to the project SupAut.
2.4.2.1 Choosing the Right Conversational AI for Project SupAut
Twilio is a cloud communications platform that provides a suite of APIs and tools for
building and managing chatbots and other messaging-based applications. It allows
developers to build chatbots and other conversational interfaces that can communicate with
users through SMS, chat apps, and other channels (Twilio, n.d.). Twilio also provides a range
of tools for managing and scaling chatbot deployments, including tools for routing and
handling customer inquiries, managing customer conversations, and tracking chatbot
performance (Twilio, n.d.).
Studies have found that Twilio is effective in enabling the development of chatbots that can
handle customer inquiries and provide timely and relevant responses in a range of industries.
For example, one study found that a chatbot developed using Twilio was able to handle a
large volume of customer inquiries and had high levels of customer satisfaction in the
financial services industry. Another study found that a chatbot developed using Twilio was
able to handle a high volume of customer inquiries and had high levels of customer
satisfaction in the airline industry (Zhang et al., 2020). A third study found that a chatbot
developed using Twilio was able to handle a high volume of customer inquiries and had
high levels of customer satisfaction in the healthcare industry (Zhang et al., 2021). These
findings suggest that Twilio is a powerful tool for building chatbots that can effectively
manage customer interactions.
In addition to its effectiveness in enabling the development of chatbots that can handle
customer inquiries, Twilio also offers a range of other features and capabilities that may be
beneficial for chatbot development. For example, Twilio provides a range of APIs and tools
for building and integrating chatbots with a variety of messaging platforms, including SMS,
WhatsApp, and Slack (Twilio, n.d.). This can be useful for reaching users through a variety
of channels and platforms and allows developers to build chatbots that can communicate with
users in real time. Twilio also provides tools for managing and scaling chatbot deployments,
including tools for routing and handling customer inquiries, managing customer
conversations, and tracking chatbot performance (Twilio, n.d.). These features can be useful
for building and deploying chatbots that can handle a large volume of customer inquiries and
provide a high level of customer service.
Twilio also offers a range of resources and support for developers building chatbots and other
conversational interfaces. For example, Twilio provides documentation, tutorials, and other
resources to help developers learn how to use its APIs and tools to build chatbots and other
messaging-based applications (Twilio, n.d.). Twilio also offers support through its developer
community forums and through its customer support team, which can assist with technical
issues and other questions related to chatbot development (Twilio, n.d.). These resources and
support can be valuable for developers looking to build and deploy chatbots using Twilio.
Overall, Twilio has demonstrated effectiveness as a tool for building and deploying chatbots
and other conversational interfaces, and has a range of capabilities and features that may be
useful for chatbot development. While the choice between Twilio and Dialog Flow will
depend on the specific needs and requirements of a given project, the studies and features
mentioned above suggest that Twilio may be a particularly effective tool for building chatbots
that can handle a large volume of customers inquiries and provide timely and relevant
responses. It is important to carefully evaluate the capabilities and features of both Twilio and
Dialog Flow in the context of a specific project to determine which tool is the best fit.
2.4.2.2 Twilio - Under the Hood
Figure 4: Initiating an outbound message and handle task creation (Twilio, 2022)
Twilio initiates an outbound message by using its API to send an HTTP POST request to the
appropriate Twilio endpoint. This request includes the phone number to which the message
should be sent, the body of the message, and other relevant details such as the sender's phone
number (Twilio, 2022).
To handle task creation on the customer response Twilio provides a variety of options. For
example, a developer can use the API to set up a webhook that will be called whenever a
customer responds to a message. This webhook could then be used to create a new task in a
task management system or trigger some other action (Twilio, 2022).
Alternatively, Twilio provides the "Gather" verb, which allows developers to collect input
from the customer via the phone keypad or voice. This input could then be used to create a
new task or trigger some other action (Twilio, 2022).
2.4.3 Selection of Transformer-based Language Model
A Transformer-based language model is essential for conducting a coherent conversation

through a chatbot. To effectively address the questions and concerns of the student, a
Transformer-based language model is superior to an RNN (Recurrent Neural Network).
Transformer models can process input sequences in parallel, while RNNs can only process
them sequentially, making them slower to train and evaluate (Vaswani et al., 2017).
Additionally, transformer models are better at capturing long-range dependencies in data and
have been demonstrated to outperform RNNs on various natural languages processing tasks,
such as language modeling and summarization (Vaswani et al., 2017). These capabilities
contribute to a more effective and efficient which is a vital requirement for the SupAut chatbot,
enabling it to understand the context of the conversation and maintain continuity. In
conclusion, for the chatbot project SupAut is seeking to build, a transformer-based language
model would be the optimal choice
2.4.3.1 Transformer-based Language Model - Under the hood
The Transformer model, introduced in the paper "Attention is All You Need" (Vaswani et al.,
2017), is a neural machine translation model that relies on self-attention mechanisms to
process input sequences and generate output sequences. It is trained to predict the next word
in a sequence given the previous words.
Figure 5: The encoder-decoder structure (Vaswani et al., 2017)
The model consists of an encoder and a decoder, both of which are made up of multiple layers
of self-attention mechanisms and feedforward neural networks (Vaswani et al., 2017). The
encoder processes the input sequence and generates a continuous representation, or context,
of the input sequence. The decoder uses this context, along with its own internal state, to
generate the output sequence (Vaswani et al., 2017).
The self-attention mechanisms in the encoder and decoder allow the model to focus on specific
parts of the input and output sequences as it processes them, rather than considering the entire
sequences at once (Vaswani et al., 2017). This allows the model to capture long-range
dependencies in the sequences and handle inputs of varying lengths (Vaswani et al., 2017).
The Transformer model is trained using supervised learning, where the goal is to predict the
correct output sequence given an input sequence and a set of target output sequences (Vaswani
et al., 2017). During training, the model is presented with input-output pairs and uses
backpropagation and gradient descent to adjust its weights and improve its predictions
(Vaswani et al., 2017).
2.4.3.2 Choosing the Right Transformer-based Language Model for Project SupAut
Transformer-based language models are a type of deep learning model that have been widely
used in natural language processing (NLP) tasks (Devlin et al., 2019; Liu et al., 2019). They
are particularly effective for tasks that involve understanding and generating text, such as
machine translation, language generation, summarization, and question answering (Devlin et
al., 2019; Liu et al., 2019). One of the key features of transformer-based language models is
their ability to process input sequences in parallel, which allows them to process long
sequences of text quickly and accurately (Vaswani et al., 2017). This is in contrast to earlier
language models, such as recurrent neural networks (RNNs), which process input sequences
sequentially and can be slow when processing long sequences (Vaswani et al., 2017).
Transformer-based language models have been shown to achieve state-of-the-art performance
on a variety of NLP tasks (Devlin et al., 2019; Liu et al., 2019). For example, the transformer-
based language model BERT (Bidirectional Encoder Representations from Transformers) has
been shown to outperform previous state-of-the-art models on a wide range of NLP tasks,
including natural language understanding, language generation, and machine translation
(Devlin et al., 2019). Other transformer-based language models, such as GPT (Generative Pre-
training Transformer) and RoBERTa (Robustly Optimized BERT), have also been shown to
achieve strong results on various NLP tasks (Radford et al., 2021; Liu et al., 2019).
GPT is a generative language model that was trained to predict the next word in a text
sequence, given the previous words. It was designed to generate human-like text and has been
used for tasks such as machine translation, language modeling, and text generation (Vaswani
et al., 2017). BERT is a transformer-based model that was trained to perform a wide range of
NLP tasks, such as question answering and natural language inference, by learning contextual
relationships between words (or subwords) in a sentence (Devlin et al., 2019). BERT is a
bidirectional model, meaning it takes into account the context both to the left and to the right
of each word, which allows it to perform well on tasks that require an understanding of the
context in which a word appears (Devlin et al., 2019)
2.4.4 Selection of Natural Language Processing (NLP) Technique
2.4.4.1 Importance of Sentiment Analysis in the Domain
Sentiment analysis is a natural language processing (NLP) technique that involves

automatically identifying and extracting subjective information from the text (Kaur &
Kautish, 2019). It is often used to determine the overall sentiment of a piece of text, such as
whether it is positive, negative, or neutral.
Sentiment analysis can be important in helping autistic kids in several ways. For example,
autistic children may have difficulty expressing their thoughts and emotions clearly and
concisely (Howlin & Blackburn, 2004). Sentiment analysis can be used to automatically
analyze the sentiment in their writing or speech, which can provide insight into their thoughts
and feelings and help caregivers to better understand and support them (Iglesias & Moreno,
2019). This can be particularly useful for children who are nonverbal or have limited
communication skills.
Autistic children may also have difficulty understanding and interpreting the emotions of
others (Howlin & Blackburn, 2004). Sentiment analysis can be used to automatically analyze
the sentiment in written or spoken communication, which can help autistic children to better
understand the emotions of others and improve their social skills (Iglesias & Moreno, 2019).
This can be especially useful for children who have difficulty interpreting facial expressions
or body language.
In addition to its potential benefits for communication and social skills, sentiment analysis can
also be useful for detecting potential mental health concerns in autistic children. Autistic
children may be more prone to mental health issues such as anxiety and depression (Howlin
& Blackburn, 2004). By using sentiment analysis to automatically monitor their
communication for signs of negative sentiment, caregivers can be alerted to potential mental
health concerns and take action to provide timely intervention (Iglesias & Moreno, 2019).
2.4.4.2 Role of Treebank Annotation in Sentiment Analysis
In sentiment analysis, treebank annotation can be used to identify the sentiment-bearing words in a
sentence and to better understand the context in which they are used through part-of-speech tagging
(Kaur & Kautish, 2019). It can also be used to identify the grammatical relationships between words
in a sentence, which can improve the accuracy of sentiment analysis by helping to better understand
the meaning and context of sentiment-bearing words through dependency parsing (Nivre, 2006).
Additionally, it can be used to disambiguate words that have multiple meanings, which can improve
the accuracy of sentiment analysis (Hung & Chen, 2016).
Treebank annotation involves adding syntactic information to a text corpus in the form of a tree
structure. There are two main approaches to treebank annotation: constituency-based, which involves
labeling the phrases in a sentence according to their syntactic function, and dependency-based, which
uses labeled arcs to represent the grammatical relationships between words in a sentence.
Figure 6 : Dependency Based Annotation (Rajput, A. 2019)
Dependency parsing (Nivre, 2006), which is a form of dependency-based annotation, can be used to
identify the grammatical relationships between words in a sentence, which can improve the accuracy
of sentiment analysis by helping to better understand the meaning and context of sentiment-bearing
words.
Figure 7: Constituency Based Annotation (Rajput, A. 2019)
Constituency-based treebank annotation (Liu et al., 2018) can also be useful for sentiment analysis,
as it can help to identify noun phrases and verb phrases, which can provide insight into the sentiment
expressed by the words in these phrases.
2.4.4.3 Choosing the Right NLP Technique for Project SupAut
The sentiment analysis technique involves analyzing the text to determine the overall
sentiment expressed in the text, such as whether it is positive, negative, or neutral. It can be
used to identify the negative or offensive language in the text. This feature can be used to
identify the negative or offensive language in the text, and its entity recognition feature to
identify specific words or phrases that do not fit with social conversing standards. Ultimately
this information can be used to rephrase the text in a way that is more appropriate and
conforms to social conversing standards.
For building a chatbot to help autistic children rephrase their sentences to fit with social
conversing standards, the Google Cloud Natural Language NLP technique has been
chosen. This technique offers several advantages in terms of the factors outlined above:
Data privacy and security: Google Cloud is a well-respected provider with a strong track
record in terms of data privacy and security (Google Cloud, 2021). They have robust
measures in place to protect sensitive data and comply with relevant laws and regulations
(Google Cloud, 2021). This makes Google Cloud Natural Language a good choice for
handling the sensitive details of the children, such as their personal information and any
medical or therapeutic information.
Integration with other tools: Google Cloud Natural Language can be easily integrated with a
variety of tools and systems, including React.JS and Twilio (Google Cloud, 2021). This means
that it can be used in conjunction with these tools to provide rephrased responses to users. The
flexibility and convenience of integration can be particularly useful for building a chatbot that
needs to work seamlessly with other systems.
Customization options: Google Cloud Natural Language offers a range of customization

options, including the ability to fine-tune the model and add specific customization options
(Google Cloud, 2021). This makes it a suitable choice for the chatbot's specific rephrasing
needs, as it allows for a high degree of flexibility and control.
Cost: While Google Cloud Natural Language may not be the most cost-effective option
available, it is likely to be a reasonable choice for a startup with limited funding (Google
Cloud, 2021). Google Cloud offers a range of pricing options and discounts that can help
make it more affordable for startups and small businesses (Google Cloud, 2021).
Technical expertise: As a team of 5 developers with intermediate expertise, the users should
be able to effectively utilize Google Cloud Natural Language without excessive difficulty
(Google Cloud, 2021). Google Cloud provides good documentation and support to help users
get started, and the interface is relatively straightforward (Google Cloud, 2021). This makes
it a suitable choice for the team's skill level.
Ease of use: Google Cloud Natural Language has a user-friendly interface and good
documentation and support, making it relatively easy to use and understand (Google Cloud,
2021). This is particularly important for a chatbot that will be used by children with autism,
as it helps to ensure that the chatbot is accessible and easy to use for all users.
Compatibility with other technology: Google Cloud Natural Language is compatible with a
wide range of technologies, including Firebase - Firestore and hosting, as well as Google
Cloud (Google Cloud, 2021). This means that it can be easily integrated with these
technologies and used in conjunction with them to build a comprehensive chatbot system.
In summary, Google Cloud Natural Language appears to be a suitable choice for building a
chatbot to help autistic children rephrase their sentences to fit with social conversing
standards. It offers strong data privacy and security measures, good integration and
customization options,
2.5 Existing Systems
2.5.1 Importance of Technological Solutions in the Domain
The use of technology, particularly software applications, in teaching children with autism
can be an important tool in supporting their learning and development. There are several
reasons for this. Firstly, children with autism often have unique learning needs that may be
better supported by technology. For example, some children with autism may benefit from
visual or auditory aids provided through technology (Anoyiannakis, 2013), and technology
can also provide opportunities for individualized instruction, which can be particularly
important for children with autism who may struggle with participation in traditional
classroom activities (Ghanouni et al., 2019). Secondly, technology can be an effective way
of engaging children with autism in the learning process. Many children with autism have an
interest in technology, and its use in the classroom can help to hold their attention and
motivate them to learn (Gómez-Marí et al., 2022). Technology can also provide opportunities
for interactive learning, which can be more engaging for children with autism than traditional,
lecture-based instruction (Anoyiannakis, 2013). Finally, technology can be a valuable tool in
helping children with autism to develop important skills such as communication, social
interaction, and problem-solving (Ghanouni et al., 2019). For example, technology can be
used to support the development of social skills through the use of social stories or video
modeling (Anoyiannakis, 2013).
2.5.2 Lack of Technological Solutions in the Domain
Several research studies explore the lack of technologies used to help teachers teach children
with autism. One study, by Anoyiannakis (2013), found that there is a limited amount of
research on the use of technology to support students with autism in the classroom. The
authors conducted a review of the literature and found that many of the studies had small
sample sizes and lacked rigorous experimental designs, making it difficult to draw strong
conclusions about the effectiveness of these technologies. This lack of research on the
effectiveness of these technologies may contribute to a lack of adoption in the classroom.
Another study, by Ghanouni et al. (2019), found that there are significant barriers to the
adoption of technology in the classroom, including a lack of funding and resources. The
authors surveyed teachers and found that cost, lack of access to technology, and lack of
training were major barriers to the use of technology in the classroom. These barriers may
prevent teachers from utilizing technologies to teach children with autism. The importance
of technology in supporting the learning and development of children with autism has been
further reinforced by the multiple instances in which the interviewees mentioned either the
unavailability of effective technological resources or their lack of knowledge regarding
potential ways to use technology to replace repetitive work (SEN Teacher, 2022; Trainee
SEN Teacher, 2022; SEN Educator Recruiter, 2022). This underscores the need for increased
access to technology and training in its use for teachers working with children with autism,
as it could potentially provide unique opportunities for engaging and motivating these
students in the learning process, as well as supporting the development of important skills
such as communication, social interaction, and problem-solving. Finally, a study by Gómez-
Marí et al (2022) found that many teachers are unaware of the potential benefits of using
technology to teach children with autism. The authors surveyed teachers and found that many
were not familiar with the various technologies available or did not know how to effectively
integrate them into their teaching practices. This lack of awareness or understanding may
also contribute to the lack of adoption of these technologies in the classroom. Overall, these
research studies support the points made about the lack of technologies used to help teachers
teach children with autism, including a lack of research on their effectiveness, lack of funding
and resources, and lack of awareness or understanding of their potential benefits.
2.5.3 Evaluation of Eolutions in the Domain
The subsequent section of this paper will focus on two significant applications that are
beneficial in helping teachers effectively utilize their time in the classroom and better
understand and prepare themselves for an enhanced teaching experience for their students.
2.5.3.1 Frontline Education
Special education management software, such as Frontline Education's offering, is a

comprehensive tool designed specifically for special education professionals (Frontline
Education, 2021). It helps to streamline and manage the complex tasks and processes
involved in special education, including student records, scheduling, assessments, and
reporting (National Center for Education Statistics, 2020). The software also provides tools
for communication, collaboration, and data analysis, making it easier for educators to track
student progress and support their academic and social-emotional needs (Frontline
Education, 2021).
According to a study published by Parsons et al., (2015), the use of special education
management software can significantly improve the efficiency and effectiveness of special
education professionals. The study found that educators who used the software reported
increased productivity, improved communication with parents and other stakeholders, and
more accurate tracking and assessment of student progress.
In addition to its benefits for educators, special education management software can also have
a positive impact on students with special needs. A study published in the McNicholl et al.,
(2019) found that students who used assistive technology, including special education
management software, demonstrated improved academic performance and social-emotional
skills.
Overall, the research supports the use of special education management software, such as
Frontline Education's offering, in improving the efficiency and effectiveness of special
education professionals and supporting the academic and social-emotional needs of
students with special needs.
2.5.3.2 Cough Drop
CoughDrop is a communication and assistive technology platform that helps individuals with
speech and language disorders to communicate more effectively (CoughDrop, 2023). The
platform features a customizable virtual communication board, text-to-speech capabilities,
and a variety of other tools and resources for individuals and professionals working with those
who have communication challenges (National Institute on Deafness and Other
Communication Disorders, 2021).
According to research conducted by the University of Kansas (2020), the use of assistive
technology such as communication boards and text-to-speech can significantly improve the
communication abilities of individuals with speech and language disorders. The study found
that individuals who used assistive technology reported increased independence, participation
in daily activities, and overall quality of life.
In addition to supporting individuals with speech and language disorders, CoughDrop also
works to support professionals in the field of speech and language therapy, providing them
with resources and support to better serve their clients (CoughDrop, 2023). A study published
Olszewski et al (2022) states that professionals who used assistive technology in their practice
reported increased efficiency and effectiveness in their work.
Overall, the research supports the work of CoughDrop in empowering individuals with
speech and language disorders to have a voice and be heard, and in supporting professionals
in their work with these individuals.
2.5.3.3 Uniqueness of the Proposed Solution
Frontline Education and CoughDrop are two technologies that excel in their respective
domains, but Project SupAut stands out due to its unique combination of these two products.
By carefully extracting the best features of both Frontline Education and CoughDrop, Project
SupAut effectively bridges the gap between management and application.
Frontline Education's well-crafted management system for student details, lesson plans, and
preparation of IEP report is an effective method of tracking student progress and has certainly
served as inspiration for the project in question. The system allows for easy organization and
access to important information, enabling teachers to efficiently utilize their time in the
classroom and better understand and prepare themselves for an enhanced teaching experience
for their students. On the other hand, CoughDrop's use of communication and assistive
technology is an excellent approach for recording and analyzing a child's communication,
social, and emotional interaction. By using this technology, teachers can gain valuable insights
into a child's development and identify areas where they may need additional support.
By combining these two applications and merging both management and raw application,
Project SupAut offers a novel approach that connects the two through additional assessment
and questions. This helps to improve both academic and other skills such as social, logical,
creative, and emotional skills. Through the integration of Frontline Education's management
system and CoughDrop's communication and assistive technology, Project SupAut offers a
comprehensive solution for teachers working with children with autism.
2.6 Chapter Summary
The main aim of this chapter was to identify the most effective approaches, concepts, and techniques
for providing a solution to the problem domain of inclusive education for students with autism. In
order to do this, the concept graph was documented as a systematic method of gathering necessary
data. The review process was then divided into three main areas: domain, technologies, and existing
work. The domain review began with an introduction to autism and a discussion of the diagnostic
criteria and prevalence of this condition. The identified problems within the domain and potential
ways to address them at a high level were also presented. The specific domain of inclusive education
for students with autism in a local setting was then explored, with a focus on approaches to inclusive
education. The final section of the domain review centered on the various challenges that teachers
face when teaching in an inclusive classroom, which is the main problem addressed in the paper. The
review of technologies involved analyzing the purpose, high-level functioning, and feature selection
of various technology tools, including AI art generation, conversational AI, transformer-based
language models, and natural language processing techniques. The advantages, limitations, usage in
projects, and alternative options for each technology were examined and compared. The review of
existing work included an analysis of Frontline Education, a special education management software,
and CoughDrop, a communication and assistive technology platform. The comparisons, advantages,
limitations, and new developments of each existing work were analyzed. Throughout the chapter, the
findings and suitability for this project were interwoven with the authors' own review and analysis.
CHAPTER 3 - Methodology
CHAPTER 3: METHODOLOGY
3.1. Chapter Overview
This chapter includes the techniques/methods used in this specific project during every phase of this
software development and how this specific project is managed. It briefly includes “What?”, “Why?”
and “When?” Specifically, the following techniques are going to be used. Also, how these methods
differ and suit this project from the researcher's perspective.
3.2. Research Methodology
3.2.1 Research Philosophy
A research philosophy is a set of beliefs or values that guides the way a researcher approaches their
work. Research philosophies can shape the types of research questions that are asked, the methods
that are used to collect and analyse data, and the way in which research findings are interpreted.
There are a few research philosophies which would guide the team in using methodologies. It is worth
noting that these philosophies are not mutually exclusive and that a researcher may draw on multiple
philosophies in their work. Some common research philosophies include empiricism, positivism,
pragmatism, pluralism, interpretivism, critical theory, constructivism etc.
Research Philosophy Key principles Focus Methods
Pragmatism Usefulness and Solving real-world Flexible, can use both

practicality; problems; practicality qualitative and
knowledge as a tool to and usefulness quantitative methods
solve practical
problems; truth
determined by
consequences
(Kaushik and Walsh,
2019)
Positivism Only knowledge Developing universal Quantitative methods,

confirmed laws and explanations; emphasis on empirical
by the sciences is objectivity data

genuine knowledge;
knowledge is gained
by gathering facts that
provide the basis for
laws (Ryan, 2018).
Interpretivism Interpretivism argues Understanding Qualitative methods,

that truth and subjective experiences such as interviews and
knowledge are and perspectives observations
subjective, as well as
culturally
and historically
situated, based on
people’s
experiences and their
understanding of them
(Ryan, 2018).
Critical Theory The object of Social justice and Qualitative methods,

study and subject of change; understanding focus on issues of
study are inextricably power and inequality social justice and
linked and that the change
researcher is always
part
of the object of
enquiry (Ryan, 2018).
Constructivism an approach that Understanding how Qualitative methods,

asserts that people reality is constructed; focus on researcher's
construct their own co-creation between role in shaping
understanding and researcher and research process
knowledge of the research participants

world through
experiencing things
and reflecting on those
experiences (Adom,
Yeboah and Ankrah,
2016).
Table 10 : Research philosophies comparison
Researchers may draw from multiple research philosophies in their work, depending on the specific
research question and context. It is important for researchers to be aware of their own research
philosophy and how it shapes their approach to research, as it can influence the results and
conclusions of their studies. The philosophy which will be followed for this project is pragmatism
since it’s the most suitable approach because of its usefulness and practicality.
3.2.2 Research Methodologies
Research methodology is the analysis of all the techniques and procedures which are applied to find,
select and evaluate research projects.
There are many different research methodologies, including qualitative, quantitative, and mixed
methods (Creswell and Creswell, 2018).
3.2.2.1 Quantitative Research
Quantitative research is a type of social science research that uses numerical data to analyse and
answer questions. It is used to measure variables, examine relationships between variables, and test
hypotheses. Quantitative research involves collecting and analyzing numerical data in order to test
hypotheses and make statistical inferences (Neuman,2018). Quantitative research is often used in the
sciences, such as psychology, sociology, economics, and medicine.
3.2.2.2 Qualitative Research
is a type of research that involves the use of qualitative data to answer questions about a given topic.
Qualitative research involves collecting and analyzing non-numerical data, such as words, images,
and sounds, in order to understand the meanings, experiences, and perspectives of individuals
(Neuman, 2018). Qualitative data is non-numerical information such as interviews, observations, and
open-ended surveys. Qualitative research is used to gain an understanding of any particular problem.
3.2.2.3 Mixed-methods Research
Mixed-methods research involves using both qualitative and quantitative methods in the same study
(Creswell and Creswell, 2018). It is used to gain a more comprehensive understanding of a research
topic than either quantitative or qualitative methods alone. Mixed methods research can be used to
answer questions about the relationships between variables, to explore and gain in-depth knowledge
3.2.2.4 Comparison between Research Methods
Qualitative Research Quantitative Research Mixed Research
Nature of data Qualitative data Quantitative data Both

collected
Forms of data Interviews, texts Numbers, Statistics Both

collected
Data Collection Observations, Surveys, experiments Both

Techniques interviews, focus
groups
Data analysis Thematic analysis, Statistical analysis Both

techniques discourse analysis, etc.
Limitations Limited May not capture the Complex design,

Generalizability, complexity and requires expertise both
subjectivity context, limited to approaches
what is measured
Strengths In-depth Generalizability, Combines strengths of

understanding of statistical power both approaches
context, rich data
Table 11 : Comparison of the research methodologies

The mixed methods research methodology will be followed throughout the project because of its own
advantages mentioned above.
3.3. Development Methodology
Software development methodology is a framework that is used to structure, plan, and control the
process of developing software (Iaeme, 2012). A development methodology is a set of principles,
practices, and tools that are used to guide the development of a product, process, or system (Boehm
et al., 2018). It defines activities performed at each step in the software development process,
including requirements gathering, design, coding, testing, deployment and maintenance. There are
various methodologies which are widely used among researchers. Such as waterfall, agile, DevOps
etc.
Out of all the methodologies available, Agile and waterfall methodologies are the most commonly
used.
The agile development methodology is a type of software development process that emphasises
collaboration, flexibility, and continuous improvement. Agile development is a set of principles for
software development that values individuals and interactions, working software, customer
collaboration, and responding to change (Agile Manifesto, 2001). It is based on iterative and
incremental development, where requirements and solutions evolve through collaboration between
self-organising cross-functional teams. Agile methodology focuses on delivering working software
quickly and frequently. It is designed to help teams respond quickly to changing customer needs and
market conditions. Agile development methodology also encourages teams to break down large
projects into smaller, more manageable pieces that can be completed in shorter time frames.
Waterfall Development is a linear, sequential approach to software development in which progress

flows in a single direction, like a waterfall. Waterfall development is a linear, sequential approach to
software development in which progress flows in a single direction, like a waterfall (Boehm et al.,
2018). It follows a defined set of steps, with each phase completed before moving on to the next. It is
more rigid and structured than agile development, with strict documentation and defined deliverables
at each phase.
Waterfall Development Agile Development
Process Linear and sequential Iterative and incremental
Documentation Heavyweight Lightweight
Requirements Fixed and defined upfront Flexible and evolving
Testing Testing after completion of Continuous testing throughout

each phase development
Risks Identifies and manages risks at Identifies and manages risks

the end early on
Changes Inflexible to changes Easily accommodate changes
Progress Measured by completion of Measured by working software

phase or deliverables
Table 12: Comparison of the development methodologies

Agile development methodology will be followed throughout the project because of its own
advantages mentioned above.
3.4 Design Methodology
Design methodology is a set of principles and processes used to guide the design of a product or
system. It is a systematic approach to problem-solving that involves understanding the problem,
gathering data, analysing the data, developing solutions, and testing the solutions. Design
methodologies are approaches to designing products, processes, or systems that involve a systematic
and structured process (Boehm et al., 2018). The goal of design methodology is to create an effective
and efficient solution that meets the needs of the user. Design methodology can be applied to any
type of product or system, from software applications to physical products.
Systems Analysis and Design Methodology (SADM) is a structured approach to the analysis, design,
implementation, and maintenance of software systems.SSAD (Structured Systems Analysis and
Design) is a methodology for designing and implementing information systems that involves a
structured and systematic approach (Yourdon & Constantine, 2016). It is based on the principles of
object-oriented programming and uses a set of techniques to analyse user requirements, design system
architecture, and develop software components. SADM is used to create systems that are reliable,
maintainable, and extensible. The methodology consists of several steps including requirements
gathering, system analysis, system design, implementation, testing and maintenance. The goal of
SADM is to create a system that meets user needs while being cost-effective and efficient.
Object-Oriented Analysis and Design (OOAD) is a software engineering methodology used for
analysing and designing an application or system through the application of object-oriented
programming. OOAD (Object-Oriented Analysis and Design) is a methodology for designing
software systems that involves modelling the system using objects and their relationships (Fowler,
2015). It focuses on the development of computer software that is based on objects, rather than
functions and logic. OOAD is a process that helps developers to analyse, design, and implement
object-oriented software systems. The methodology consists of several steps, including object
identification, class definition, relationship definition, interface definition, and implementation. The
goal of OOAD is to create a system that is easy to maintain and extend over time.
Structured Systems Analysis Object-Oriented Analysis and

and Design (SSAD) Design (OOAD)
Approach Top-down, waterfall Iterative, incremental
Focus of analysis and design System-level Object-level
Techniques Used Data flow diagrams, entity- Use case diagrams, class
relationship diagrams, diagrams, object-oriented
structured design techniques design techniques
Limitations Inflexible to changing Requires more time and

requirements, difficult to resources upfront, requires a
incorporate new technologies deeper understanding of
object-oriented concepts
Benefits Clearly defined process, easy Reusable components,

to understand and document flexibility to change
requirements
Table 13 : Comparison of design methodologies

According to the above comparisons and details, OOAD will be used as the design methodology for
this specific project.
3.5 Evaluation Methodology
Evaluation methodologies are approaches or frameworks used to assess the effectiveness, efficiency,
and impact of a program, policy, or project. The evaluation methodology is the process of collecting
and analysing data in order to assess the effectiveness, efficiency, or impact of a program,
intervention, or policy (Patton, 2016). There are many different evaluation methodologies, including
1. Experimental designs, in which a treatment group receives the intervention being evaluated and a
control group does not, and the outcomes of the two groups are compared.
2. Quasi-experimental designs, in which a treatment group and a control group are not randomly
assigned, but the groups are still compared to assess the impact of the intervention.
3. Descriptive or process evaluation, which focuses on understanding and describing the
characteristics of a program or the processes by which it was implemented.
4. Outcomes-based evaluation, which assesses the impact of a program or intervention on the
outcomes or outcomes it is intended to influence.
5. Cost-benefit analysis, which compares the costs of a program to the benefits it generates in order
to determine its overall economic efficiency.
6. Impact evaluation, which assesses the long-term effects of a program or intervention on the
individuals or systems it is intended to benefit.
The choice of evaluation methodology depends on the goals and objectives of the evaluation, the
resources available, and the context in which the program is being implemented.
Outcomes-based evaluations look great and it suits better than the other evaluation methodologies
available. And here’s why,
Outcome-Based Evaluation Other Types of Evaluation
Focus Results or outcomes Processes or activities
Data Collection During and after During
Comparison Point Baseline measures No comparison point
Uses Assessing effectiveness Improving process
Examples Impact evaluations, Formative evaluations,

effectiveness evaluation summative evaluations
Table 14 : Comparison of evaluation methodologies
Outcome-based evaluation is focused on measuring the results or outcomes of a program, policy, or

intervention. In contrast, other types of evaluation may focus on the processes or activities involved
in implementing an intervention.
Outcome-based evaluation involves collecting data during and after the implementation of an
intervention, in order to measure the impact on the targeted outcomes. Other types of evaluation may
only involve collecting data during the implementation of an intervention.
Outcome-based evaluation includes a comparison point in the form of baseline measures, which
provide a reference point for evaluating the impact of the intervention. Other types of evaluation may
not include a comparison point.
Outcome-based evaluation is typically used to assess the effectiveness of an intervention, while other
types of evaluation may be used to improve processes or identify areas for improvement.
Examples of other types of evaluation include formative evaluations, which are conducted during the
planning or development phase of an intervention, and summative evaluations, which are conducted
at the end of an intervention to assess overall effectiveness.
3.5. Project Management Methodology
In simply stated, project management methodology is the process of using skills, information, tools,
methods, and experience to accomplish project goals and objectives. In other words, it is the
organization and management of resources to present a project. If all goals and requirements are met
within the specified time frame and price range, the design can be deemed successful. Project
management is therefore crucial to the project's success.
Numerous project management methodologies exist. There are several of them, including Slack,
ClickUp, Scrum, Kanban, Scrumban, Lean, eXtreme Programming (XP), Waterfall, PRINCE2, and
Six Sigma.
Since the authors for this project work from home, it was important to keep track of their research
output as well as to tell them of the big picture and the deadlines that were of the utmost importance.
For the writers' and the project's well-being, it was crucial that they stay in touch often. (Tomlinson,
2020).
The authors also want software that works on both desktop and mobile platforms so they can stay
connected and organized wherever they are. Slack and ClickUp has been chosen as the technique for
project management because of factors and the project types. This is a bettered and reviewed
interpretation of slack methodology. It’s an extensively conceded and practised methodology
worldwide.
3.6 Team Work Breakdown Structure (WBS)
Each report's tasks were distributed among the team members according to their skills, preferences,
and the objectives that had to be achieved. The group members are expected to research and gather
information on the assigned topics, present it to the sessions, and address any questions about the
topic. Following the member discussion, the pertinent document is amended and finished.
For better clarity, the WBS used for the project SupAut has been moved to the Appendix section.
For a more understandable form of the WBS, see Appendix - Section B.1.
3.7 Gantt Chart
A Gantt chart is a commonly used graphical depiction of a project schedule. It's a type of bar chart
showing the start and finish dates of a project's elements such as resources, planning, and
dependencies.
3.7.1 Gantt Chart Diagram

The Gantt Chart Diagram structure has been moved to Appendix - Section B.2 for better clarity.
3.8 Usage Of Project Management And Collaboration Software In The Project
3.8.1 ClickUp
It is a crucial task and project management tool that might help to simplify and enhance procedures.
It can plan tasks, keep an eye on modifications, control teams, and keep tabs on how much time is
spent on each activity. Additionally, it enables users to change views and see things from various
angles, both of which are essential qualities. In contrast to other applications, ClickUp offers several
essential features.
Figure 10: Task breakdowns for each group member in ClickUp.
Figure 11: Snip of Gantt chart usage in ClickUp.
3.8.2 Slack
It is software made to let companies interact discreetly or in groups using direct messaging. Users
may continue working without providing context for each link or document supplied if all pertinent
papers, chats, and options are available in one place.
Figure 12: Slack application for group collaboration and effective communication
3.8.3 Google Meet
Google's video conferencing service is known as Google Meet. It doesn't have too many functions
and is simple to use. Anyone with a Google Account has the ability to set up an online meeting. Users
who want to work across platforms can still have an infinite number of meetings. Additionally, users
may join meetings straight from Gmail or Google Calendar, so there is no need to download an app.
Figure 13: Snapshot of a Google Meet session
Figure 14: Snip of a video conference using Google Meet
3.9 Risks and Mitigation
No Risk Item and Description Severity Frequency Mitigation Plan
R1 Lack of Knowledge: High High Requesting/getting

As this project includes guidance from the
Machine Learning techniques experts and constant
and other technical components self-studying in the
it is tough to comprehend at relevant domain.
times.
R2 Evolving Requirements Medium High Keep track of the

affection in Scope: scope of the project.
During the project, the whenever new
stakeholders would want to requirements need to
change the requirements, which be added and
would make delays and prioritize the
problems in the development. requirements.
R3 Troubles in Keeping up with High High Breaking down the

deadlines: work of members
Insufficiency of time to learn and Setting personal

and work accordingly with deadlines for each
research, designing, task. Getting help
implementations, testing etc. from a member and
being accountable
R4 Troubles During High High Sort the

Implementation requirements in the
preferred order of
stakeholders and
work according to it
R5 Wrong technologies: High High Requesting help

Chosen Technologies and from the Experts and
methods for the project working accordingly
(Algorithms, toolkits, etc )
either doesn’t suit or work.
R6 Personal Issues of Co- Low High Being prepared for

workers: situations like this
Personal Problems faced by and pre-planning the
team members deviating them work ahead.
from the assigned work which
will affect the development.
Table 15 : Risks and Mitigation
All the methodologies which were used for this specific project are included in this chapter. The
chapter began with an introduction. Eventually, the methodologies which were used for every phase
of the project (research, development, design, evaluation and project management) were explained.
Each methodology had its own Introduction and what are the widely or commonly used
methodologies in the specific phase. Also, it included which one (methodology) is better and more
relevant to that specific phase of the project. Mixed methods research was chosen for the research
methodology, Object-oriented design was chosen as the design methodology. Agile Development
was chosen as the Development methodology and Outcomes-based evaluation was chosen as well. In
the end, the risks and mitigation have been mentioned including the most suitable mitigation plan.
To ensure that the project team and all stakeholders are in agreement on how the project will be
managed, it is essential to be able to manage the project rigorously and proactively. It is strongly
recommended to concentrate on project management while maintaining the format and content in
mind because it is the foundation of the entire project and is essential to its success.
The project's Gantt chart, workgroup breakdown chart, and project management and collaboration
software usage are all shown in this chapter. The assessment of the project's risks and their mitigation
has been emphasized, along with an appropriate mitigation strategy. The following chapter is System
Requirements Specification, which aids in determining the characteristics and behaviours of the
system.
CHAPTER 4 - System Requirement Specification
CHAPTER 4: SYSTEM REQUIREMENTS SPECIFICATION (SRS)
The previous chapter mainly concentrated on the many approaches that may be taken to finish Project
SupAut. This chapter will start out with a stakeholder analysis. In an onion model, every stakeholder
in the SupAut system will be represented, along with a description of how they will influence SupAut.
The analysis of the information received from the questionnaires and interviews was then reviewed,
followed by a discussion of requirement elicitation approaches. This chapter also includes other
analytical models, including use case diagrams, context diagrams, and domain model diagrams.
Finally, SupAut's functional and non-functional criteria have been outlined in detail along with their
relative importance to the system.
4.2 Stakeholder Analysis
4.2.1 Onion Model
Figure 15: The Onion Model (Self Composed)
4.2.2 Stakeholder Descriptions
Stakeholder Viewpoint
Functional beneficiary
Autistic Kids Kids who communicate teacher through the chatbot
Teachers who teach autistic Teachers who communicate kid through the chatbot and the
kids check progress of the autistic kid
Consultant Therapists People who check medical treatment level and diagnose the
progress of the autistic kid
Financial beneficiary
SEN Unit Special Education Unit of the school which gets funds and
equipments by NGO groups and Caregivers
School Management Gets funds by caregivers and Old Pupils’ Association
Social beneficiary
Caregiver Parents of autistic kids who can support their children to get a
better improvement of education with the support of ChatBot.
School Management Supply more facilities according to the improvements of ASD

kids to perform the results of Special Needs Children’s Unit
Ministry of Education Create different kind of lesson plans for ASD kids that can get
support through the ChatBot
Public People who can interact with ASD kids who have the general
knowledge as a normal kid.
Operational beneficiary
System Administrators Involve giving routine commands and monitoring

outputs from the website
Data Scientists Involve advising developers and system administrators about

how to operate the ChatBot.
Negative Stakeholders
Assistant Teachers Assistant teachers may be get out of their job role because of the
SupAut implementation
Regulatory
Developer Responsible for regulating the quality, safety or other aspects of

the website
Experts
Teachers who teach autistic Provides information about ASD kids and their lesson plans to
kids the developer.
Table 16: The Stakeholder descriptions table
4.3 Selection of Requirement Elicitation Techniques/Methods
The many strategies known as requirement elicitation techniques are used to collect requirements
from users and non-users to varying degrees. The requirement elicitation approaches were chosen
during the research phase so that they could be used to apply them and collect requirements from the
stakeholders of our system. This section discusses the benefits and cons of the approaches literature
review, surveys, interviews, brainstorming, and prototyping.
4.3.1 Literature Review
One of the most important methods used during the requirement elicitation phase to obtain knowledge
about the domain, the technologies that must be used, and the past work in the domain is literature
study. An in-depth and comprehensive assessment of the current system is provided through a
literature review.
Advantages Disadvantages
A literature review will provide a more A literature review will take more time since a
comprehensive understanding of the field, lot of research should be conducted and all the
which is preferable to learning about the field research needs to be reviewed.
through questionnaires and interviews with
domain experts. Because only the domain
expert's perspective may be presented in
interviews or questionnaires.
Gives a brief overview of the technologies that

are now in use in the field and those that are
still being developed.
Through a literature review the different Since the most recent technology is frequently
approaches that can be used to implement added, the literature review is a requirement
SuppAut could be found. elicitation technique that needs to be updated
daily. The most crucial characteristics of the
area will be overlooked, and the most recent
research in the field won't be available.
Findings The team gained a wider understanding of the

area as well as knowledge of the technologies
already in use in the domain and those that
have not yet been put into practice there. The
many methods that can be employed to
implement SupAut were also discovered
through the literature review.
Table 17 : Advantages,Disadvantages and findings of the literature review
4.3.2 Questionnaire
A questionnaire was prepared using Google Forms and distributed to the target audience via email
and social media in order to ascertain their needs. The benefits and drawbacks of sending an online
survey are covered here.
Because they will be the system's future Poor response rate as some users of the target
users, the needs collected through audience do not contribute to the survey.
questionnaires from the target population
are very significant and helpful.
Saves time because a larger audience can The participants put in different efforts to answer
be reached. the questions so the accuracy of the results is
affected.
With the analysis features already built into It can be challenging to analyse the data and draw
the Google Forms application, the results conclusions because participants can give a
could be easily debated and analysed. variety of replies to the questions.
Findings We found that questionnaires helped us gain a

broad understanding of the public's perception of
our system. The team's main findings included the
prevalence of autism in society, the need for a
system like SupAut, and the respondents'
perceptions of the qualities of SupAut.
Table 18 :Advantages,Disadvantages and findings of the literature review
4.3.3 Interviews
Interviews are conducted mainly with domain experts in order to identify the current requirements of
the system.
The entire team will get insight into the user's A large audience cannot be reached so
perspective and will then be able to modify the validation of the system will not be done.
system to meet the needs of the user.
The ability to get direct feedback from users The needs given by the interviewee can be
on how our system should work is a major from his or her perspective and won't
advantage of interviews. Follow-up questions necessarily be requirements for a large
might be asked and a problem could be audience.

resolved right away.
Findings Interviews were conducted with subject matter

specialists, particularly special needs teachers.
The team mainly gained insight into the
specifications that the system required to meet
and the challenges that the community of
special needs teachers now faces. Additionally,
the team used these interviews to polish the
system's conceptual framework.
Table 19 :Advantages,Disadvantages and findings of the interviews
4.3.4 Brainstorming
The team engaged in brainstorming during every stage of the project. Both team-wise and mentor-led
brainstorming sessions were held. When choosing fresh ideas and determining the needs of the
system, the brainstorming technique was crucial at the beginning of the project.
New system requirements may be discovered Receiving conflicting requests from members
as a result of brainstorming, and new concepts of the team. Consequently, it would be
for the system's view, its technology, and its challenging to determine the system's ultimate
functionalities could be discussed. requirements.
The final system would be a developed The team will constantly changing ideas as a
version of the first system because team result of the lengthy process of brainstorming.
members have different points of view.
Findings Through brainstorming, the team selected the

SupAut system's features. Brainstorming was
utilised to determine the methodology,
technologies employed, requirements, final
outlook of the system, and all other key
elements of the SupAut system.
Table 20 :Advantages,Disadvantages and findings of brainstorming
4.3.5 Prototyping
Prototyping is creating a model of the system which is similar to the final output.
Making a system that is similar to the one that It takes time to create the final prototype.
will be used in the end will make it easier for
the system's stakeholders to understand the
requirements that must be met.
The team members will also have a clear idea Some stakeholders may not like different
of the final system so that no errors would be features of the system so it would be difficult
done in the implementation stage. to proceed into the implementation of the real
system.
Findings The team also developed a clear vision for the

design of the ultimate system. Both the
additional features that should be added to the
system and the technologies that should be
used to achieve the desired system were
identified. The final SupAut system was made
obvious to the external parties as well.
Table 21 :Advantages,Disadvantages and findings of Prototyping
4.4 Discussion/ Analysis of Results
4.4.1 Questionnaire Findings
The Questionnaire Findings table has been relocated to the Appendix, Section C.2, along with the
attached screenshot for the survey form can be found in Appendix, Section C.1, for improved clarity.
Please refer to that section for a detailed version of the diagram.
4.4.2 Interview Findings
The Interview Findings table conducted for the purpose of the research for has been relocated to the
Appendix, Section C.3, for improved clarity. Please refer to that section for a detailed version of the
diagram.
4.4.3 Literature Review Findings
The results of the literature review were instrumental in shaping the requirements for this project. The
decisions and choices made in regards to the major components of the project were informed by the
literature review, which allowed the team to gain a comprehensive understanding of the proposed
solution. This preparation was crucial, as the project was limited to a three-month timeline. In order
to meet this constraint, the team decided to utilize various methods, techniques, and frameworks in
the development of the project making an informed selection of the right AI Art Generator,
Conversational AI, Transformer-based language model, and Natural Language Processing (NLP)
techniques. The thorough research conducted during the literature review phase enabled the team to
make informed choices and move forward with confidence in the implementation stage.
The tabulazied literature review findings table has been relocated to the Appendix, Section C.4, for
improved clarity. Please refer to that section for a detailed version of the diagram.
4.4.4 Summary of Findings
Findings LR Survey Brainstorming Interviews Prototyping

LR
Should be a common ✔ ✔ ✔ ✔ ✔
application for teachers,
students, and caregivers.
Should be able to help teachers ✔ ✔ ✔ ✔

to tutor children
Should facilitate the children ✔ ✔ ✔

to learn their subject material
in a more effective way.
Identification of the most ✔ ✔ ✔ ✔

important features that should
be included in SupAut.
Should be able to visualize the ✔ ✔ ✔

question.
Should be able to break down ✔ ✔ ✔
the question.
Should be able to rephrase the ✔ ✔ ✔

answer/question.
Should have a chatbot. ✔ ✔ ✔ ✔ ✔
4.5 Context Diagram
The proposed system's boundaries and interactions with the indicated internal and external
components are defined in the context diagram. Additionally, it displays an overview of the data flow
between the system and the outside world.
Figure 16: The Context Diagram (Self Composed)
4.6 Use Case Diagram
A straightforward graphical description of user interactions within a system is provided by the use
case diagram. The principal party interacting with the system is the teacher. The student, who interacts
more with the chatbot UI than the dashboard UI, is the supporting actor.
Figure 34 :Use case diagram(Self Composed)
4.6.1 Use Case Descriptions
Use Case Name Ask questions
Use Case ID UC - 002
Description The teacher asks a question(a task to be done) from the student.
Priority High
Primary Actor Teacher
Supporting actors Student
Pre-Conditions The student contact number should be in the teacher’s list of students in
the class and also the student should have the teachers number.
Trigger The question will appear in the WhatsApp mobile app/WhatsApp desktop
version of the student.
Main flow Actors System
1. The teacher will ask a 2. The system will display the

question/questions question/questions that should be
from the student. attempted by the student.
Exception flow Actors System
None None
Alternate flow Actors System
None None
Exclusions None
Post Conditions The student will see the question or questions sent by the teacher in his
WhatsApp desktop/mobile app. The dashboard will keep track of the chat
activities.
Table 28 :Use Case descriptions for “Ask Questions”
Use Case Name Visualise
Description In this scenario, a student approaches the teacher with a question (or a
task to do), which the chatbot then forwards to the student. When a
middle school student with autism requests it or when there is evidence
of a learning disability on their part, the chatbot can visualize the
question the teacher is asking in a form that the student can understand.
Priority High
Primary Actor student
Supporting Actors teacher
Pre-Conditions The student's contact information should be included in the teacher's list
of class members, along with any assigned questions that the student still
needs to attempt.
Trigger The chatbot will provide a visual image related to the question(Visualise
the question).
1. Student triggers the use of 3. The question will be forwarded.

the chatbot, to attempt to 4. Upon the student's request, the system
answer assigned questions. will display an image or photos that are
2. The learner will indicate pertinent to the subject.
that they are having trouble
understanding questions and
will ask for visual assistance.
When the teacher delivers a Instead of moving forward with question

message, the autistic student visualization, the system will let the
can reply if they don't need the child respond.
visualization function.
None None
Exclusions Visualise
Post Conditions The student can view the related visual image from his WhatsApp.The
system will record the chat activity in the dashboard.
Table 29 : Use Case descriptions for “Visualize”
Use Case Name Breakdown
Description In this use scenario, a student approaches the teacher with a question (or a
task to do), which the chatbot then forwards to the student. When a teacher
asks a question from a student, the chatbot can break it down so that the
student can grasp it, either at the student's request or when there is evidence
that the student is having trouble understanding.
Priority High
Primary Actor child
Pre-Conditions The teacher's list of pupils should include the student's phone number, and
questions that the student hasn't tried yet should be assigned.
Trigger The chatbot will display the question which is broken down into
understandable parts.
1. Student triggers the use of 3. The question will be forwarded.

the chatbot, to attempt to
answer assigned questions. 4. The system will provide a step-by-step
2. The pupil will express analysis of the question and deliver it in a
frustration about having way that the autistic youngster can
trouble understanding the understand.
question and will ask for it to
be broken down.
When the teacher sends a Instead of continuing to break down the

message, the autistic student question, the algorithm will let the child
can reply if they don't need the respond.
breaking down feature.
None None
Exclusions Breaking down of the question.
Post Conditions The student can view the broken-down question from his/her WhatsApp.
The system will record the chat activity in the dashboard.
Table 30 :Use Case descriptions for “Breakdown”
Use Case Name Rephrase
Description In this use case, if a student submits a statement that goes against social
conversational standards during the course of the conversation, the chatbot
can reword it in line with those norms.
Priority High
Primary Actor child
Pre-Conditions The teacher's list of pupils should include the student's phone number, and
questions that the student hasn't tried yet should be assigned.
Trigger The chatbot will show the rephrased question.
1. The chatbot is activated by 3. The question will be forwarded.

the student to try to respond to 4. The statement will be rephrased by the
the queries given to it. algorithm to reflect social talking norms
2. A statement that the student before being forwarded to the teacher.
types will be rephrased and
forwarded to the middle school
instructor if it is in conflict
with social chatting norms.
When the teacher sends a Instead of continuing to break down the

message, the autistic student question, the algorithm will let the
can reply if they don't need the student respond.
breaking down feature.
None None
Exclusions Rephrasing
Post Conditions The student's revised inquiries and responses will be visible to the teacher.
The dashboard will have a record of the conversation activity.
Table 31 :Use Case descriptions for “Rephrasing”
Use Case Name View IEP Report
Description This use case sends a monthly report to the caregiver and the teacher about
the progress of the student.
Priority High
Primary Actor teacher
Supporting Actors student, caregiver
Pre-Conditions Using the appropriate credentials, the teacher should be logged into the
system. The user and the autistic student should both have the other's
contact information. The student should have at least attempted one
question and the caregiver information should be provided.
Trigger The teacher can click the send IEP report button or the download IEP
report button
1. The teacher logins into 4. Auto generates the progress

the dashboard. report considering the chatbot's
2. Navigates to the correct behavior.
student on the page.
3. He/She can download 5. The automatically generated

the IEP report or send progress chart will provide a
the IEP report to the summary of the pupils' social,
caregiver. visual,logical, time-management,
and creative abilities. For
instance, the learner will make
progress if they can
independently grasp the questions
and answers.
Non None
Alternate flow Aectors System
None None
Exclusions None
Post Conditions The teacher will have access to the progress chart locally, and using the
transmit IEP report option, she can email it to the caregiver.
Table 33 :Use Case descriptions for viewing IEP report
4.7 Domain Model
Figure 18: The Domain Model diagram (Self Composed)
4.8 Functional Requirements (with prioritization)
Critical – The requirements that are critically needed in the successful completion
Desirable – The requirements that can add value, but are not required immediately
Luxury – The requirements that would add luxury to the system
Requirements list Priority Description

Level
FR1 Teacher should be able to Critical It is required that the middle school teacher
authenticate authenticates to the system using the correct
password.
FR2 Teacher should be able to ask Critical It is required that the middle school teacher
a question. asks a question from the student.
FR3 The question should be Desirable The student should be able to get a visual aid
visualized. for the question on the request of him/her.
FR4 The question should be Desirable The student should be able to obtain a
broken down into broken down version of the question on the
understandable parts. request of him/her.
FR5 The question/answer should Desirable The middle school teacher should be able to
be rephrased according to the view a rephrased answer that is sent by the
social conversing norms. student.
FR6 The student should be able to Critical It is required that the student posts a
post an answer/question question/answer back to the teacher.
FR7 Creating a monthly IEP report Critical It is necessary to evaluate how well kids
responded to the questions provided by the
teachers in order to compile a monthly IEP
report. When creating the IEP report, the use
of the features visualisation, breaking down,
and rephrasing will also be taken into
account.
Table 34 :The Functional Requirements table.
4.9 Non-Functional Requirements
Non-Functional Priority Level Description

Requirements
Usability Critical The application's usability is crucial. Both the chatbot and the
instructor dashboard should be easy to use. Both the teacher
and the student should benefit from the system saving time.
Extensibility Desirable The system ought to be changed so that new features may be
introduced without impairing the system's core functionality.
Reliability Critical Both the teacher and the student should be able to rely on the
system. The pupils must obtain the proper visual image(s) or
the proper broken down form of the question through
visualisation and breaking down the inquiry. Similar to this,
the teacher must receive the student's correctly formulated
query, and the IEP report must be produced in a trustworthy
manner utilising statistics because it affects the student's
future.
Performance Critical Every time the system is utilised, it should operate as

expected. There should be no errors in the system.
Accuracy Critical The results of visualisation,breaking down and rephrasing

should be accurate. The IEP report should be accurate. Since
this a system which is related to education over 95% accuracy
is expected.
Table 35 :The Non Functional Requirements table
The identification of the system's stakeholders served as the beginning of this chapter. An onion
model was used to represent each stakeholder and describe how they affected the system. The analysis
of the various requirement elicitation approaches was followed by the depiction of all the findings,
followed by the use case diagram, use case descriptions, context diagram, and domain model diagram
for the Project SupAut. Finally, the requirements—both functional and nonfunctional—were stated
along with their relative priorities
CHAPTER 5 - Social, Legal, Ethical and Professional Issues
CHAPTER 5: SOCIAL, LEGAL, ETHICAL AND PROFESSIONAL

ISSUES
It is important to take the required precautions as the project progresses in order to avoid any legal,
ethical, and professional problems. This chapter goes into great length on the consequences and the
strategy the team uses to deal with these problems. It aims to address every issue using the SLEP
analysis. The project's impact on society, law, and ethics, and it is also made apparent how these
issues relate to the BCS code of conduct as well as any beneficial or bad impacts on the three parties
stated above, are discussed here. Also, this chapter discusses mitigating strategies for solving these
problems.
5.2. SLEP Issues and Mitigation
Maintaining a clear head on social, legal, ethical, and professional matters is crucial throughout the
SDGP. Since the start of this undertaking, every ethical boundary has been researched and respected.
In this part, each of these facts is given and discussed. Working on this aspect of the project involves
all team members. The social, legal, ethical, and professional issues with the mitigations related to
BCS code of conduct of the Project SupAut are detailed below.
5.2.1 Social Issues
When considering social issues, challenges with the project are minimal. This project will not have a
cultural impact because it does not require a background in any religion, politics, or ethnicity.
Moreover, this program won't lead to any racist or extremist problems.
The SupAut initiative has minimal impact on society and culture. The chatbot system will have an
impact on non-native English speakers because it was primarily developed in English. A subsequent
update will include support for multilingual languages. The online application may connect with users
directly from their homes, which can help students who struggle with social isolation and isolation.
5.2.2 Legal Issues
The author has ensured that this application is created in accordance with data protection laws. In
order to protect user privacy, survey data will be compiled. The software used was all free and open
source, and it also contained applications that the university had authorised for distribution. Primarily
educational goals will be served by the gathered information.
Data protection rules were given the most focus during the project's development. Dall.E a state-of-the-art
AI art generator that has ability to generate unique and creative artworks has been used in this project
and Dall.E has a large and diverse datasets over 50million images and texts these data sets has been
used within the terms and regulations. Where software utilized in this project was legally acquired,
such as Twilio and Google cloud Natural Language, software licensing is also given top emphasis.
Twilio is a cloud communications platform that provides a suite of APIs and tools for building and
managing chatbots and other messaging-based applications and NLP models that have been used
throughout the project these particular platforms are not open source, and credits for the use of the
softwares were gained through participation of Hackthons. Other project management applications,
like Slack and ClickUp, only have community versions. GitHub, a well-known open source platform,
has been utilized to keep our project repository and to maintain the development process within the
team.
The papers used for the LR were carefully chosen from peer-reviewed prior studies, and as part of
the study, they were copyrighted and cited in accordance with the Harvard Westminster referencing
style. While the collected data through Interviews and Questionnaires were tested anonymously and
the respondents' privacy was respected, no personal information from the respondents was stored in
the questionnaires sent throughout the requirement gathering stage.
5.2.3 Ethical Issues
The respondents were asked for their consent to use their email address to represent their identify at
the beginning of the SupAut Project survey. In the SRS chapter, the survey responses are examined.
Personal information (such as email addresses) was used to confirm the users' identities, and it is
explicitly stated in the survey that no information would be used for unrelated reasons unless it is
pertinent to the question. Thus, the privacy of usage is guaranteed.
The domain experts who conducted the interviews were informed of the purpose of the interviews
and made clear how it will be used in the project.
The SupAut project will only use the data and information it collects; none of it will be made available
to outside parties. Aside from making sure that none of the software utilized in the project has been
cracked or pirated, it is being legitimately produced.
5.2.4 Professional Issues
This system was created with great professionalism. Everyone who participated in the data collection
process and the Interviews and questionnaire was aware of the job they were doing and their identity
was kept anonymous throughout the study. The privacy and confidentiality of user information is
intended. Project SupAut's creators made the required preparations to guarantee that the interviewees,
questioners, and data were handled properly.
Each image that was submitted to the report had a citation, and all sources utilized for the project
were properly cited in the references section.
5.3 Chapter Summary
In this chapter, social, legal, ethical, and professional issues relating to this project were taken into
consideration, recognized, and handled in line with the BCS code of conduct both during the
implementation of the program. We have reviewed the procedures that must be taken to reduce the
SLEP issues that are related to our project in order to prevent any unfavorable effects from the external
environment.
CHAPTER 6 - System Architecture & Design
CHAPTER 6: SYSTEM ARCHITECTURE & DESIGN

The system architectural design and diagrams used to create Project SupAut are primarily covered in
this chapter. This chapter includes the system's high level architecture design, the class diagram, the
sequence diagram, the activity diagram, and the UI wireframes.
6.2 System Architecture Design
The high level architecture diagram for Project SupAut is shown in the picture below. The structure
of the diagram consists of five layers: frontend, Conversational AI, API, NLP library, and database.
From the perspective of the project's developers, the suggested system will begin with the frontend,
move on to the Conversational AI, API, and NLP library, and then finish with the database layer.
The user will initially engage with the chatbot and dashboard user interfaces. The system will then
go on to conversational AI to accomplish the tasks of student identification, sentiment detection for
antisocial behaviour, and comprehension difficulties detection. After that, the system will continue
on to the API and NLP libraries, respectively, to rephrase phrases, visualise elements, deconstruct
instructions, and ultimately assess student abilities.Finally the system will move onto the database
layer to store question details and student details.
Figure
19:The High Level Architecture Diagram (Self Composed)
6.3 System Design
6.3.1 Class Diagram
The relationships between the primary classes in Project SupAut are depicted in the class diagrams.
The identified classes each had a significant role in the proposed system and a distinct set of properties
and methods. The diagram effectively maps each of SupAut's structural elements, as well as their
behaviours and characteristics.
Figure 20 : The Class Diagram (Self Composed)
Class Responsibility Collaboration
Dashboard Contains the important details of the Teacher,Chatbot,Dashboard

SupAut system. page,Student page,Question
Chatbot Facilitates the chatbot activity. Dashboard,Analyse
skill,Student,Question
DashboardPage Contains the student progress graph and the Dashboard, Question
student count graph with add student and
add question functionalities.
Question Contains the details of the question. Dashboard,Student,Teacher
AnalyseSkills Evaluates the different skills of the Chatbot

students.
Teacher Contains the details of the teacher Dashboard,Question
Student Contains the details of the students. Question,Chatbot,Student page
Student Page Gives an extended overview of the student Dashboard,student

details and student progress.
Table 37 : Descriptions of classes in Class Diagrams
6.3.2 Sequence Diagram
6.3.2.1 Sequence diagram for the dashboard
Figure 21: Sequence diagram for the dashboard(Self Composed)

The actions of the dashboard are shown in the above sequence diagram. The question and the student
information can both be added by the teacher. The student information and the question information
will be stored in the database. As a result, the dashboard in the database will display student and
question information so that the teacher may see it beforehand. The IEP report will then be prepared
by the API and displayed on the dashboard. The IEP report is then accessible to the user.
6.3.2.2 Sequence diagram for the Chatbot
Figure 22:Sequence diagram for the Chatbot(Self Composed)

Above is a picture of the chatbot's flowchart. When a user submits a query, the SupAut system
evaluates it. The technology picks up on any unfavourable tone and comprehension issues, and the
API then analyses the data and simplifies the instructions. The database receives the query and
forwards it to the chatbot interface so the student may see it. The student responds via the chatbot,
and after being rephrased and skill-analysed by the API, the answer is sent to the database.
6.3.3 UI Design
For the SupAut application, wireframes and design prototypes were created prior to the
implementation phase. The foundational pages that will be provided by the system are shown below.
6.3.3.1 Dashboard page
The dashboard will show the number of users, newly hired employees, new questions added overall,
the number of modules, and the average amount of time spent on the dashboard. There is also a line
graph showing student progress and a bar chart showing the number of students. The middle school
teacher can use this page to add a kid or a query.
Figure 23 : The Dashboard page
6.3.3.2 Add student page
The teacher will be taken to the add student page if they click the add student button on the dashboard
page. The add student page has fields for the student's name, student contact number, caregiver's
name, caregiver contact number, and caregiver email address. A teacher can add a student to the
SupAut system by clicking the add student button after all these forms have been filled in.
Figure 24 :The Add student page
6.3.3.3 Add question page
The teacher will be taken to the add question page if they click the add question button on the
dashboard page. On the website, there are fields for adding the grade, subject, question, description,
and the sample answer. The middle school teacher can click the save button to save the question data
once all of these fields have been filled in.
Figure 25 :The Add question page
6.3.3.4 Student Details page
A picture of the student will be included on the student details page. There are fields for the student's
name, student contact information, caregiver's name, caregiver contact information, and caregiver
email address. The student page contains a progress graph specific to the student. The visual, logical,
creative, and time management abilities of the specific autistic child whose profile this is shown in
this progress chart. The chart will be made based on how the pupils responded to the questions. The
middle school teacher can send the IEP report to the carer's email by simply clicking the Send IEP
report button. If the teacher wishes to download the student's IEP report, she can do so by selecting
the button Download IEP report.
Figure 26 :The student details page
6.3.3.5 Question page
The question page shows the grade for which the question is designated, the question's number, the
question, and the question's responses.
Figure 27 :The Question page
6.3.4 Process Flow Chart
The user and system-performed actions are depicted in the process activity diagram below. The
system's functions and their flow are summarised in this activity diagram.The concept graph can be
found in the appendix - Section D - Design.
6.4. Chapter Summary
The high level system architecture diagram serves as the chapter's starting point before moving on to
the low level system diagrams, which include the class diagram, sequence diagram, and activity
diagram. This chapter concludes with the UI designs and wireframes of the Project SupAut. All of
these diagrams come with a thorough explanation.
References
REFERENCES
Able, H. et al. (2014). Views from the Trenches. Teacher Education and Special Education: The
Journal of the Teacher Education Division of the Council for Exceptional Children, 38 (1), 44–57.
Available from https://doi.org/10.1177/0888406414558096 [Accessed 9 January 2023].
Adom, D., Yeboah, A. and Ankrah, A. (2016). CONSTRUCTIVISM PHILOSOPHICAL

PARADIGM: IMPLICATION FOR RESEARCH, TEACHING AND LEARNING. Global Journal
of Arts Humanities and Social Sciences, 4 (10), 1–9. Available from
https://www.academia.edu/download/60883069/Constructivism-Philosophical-Paradigm-
Implication-for-Research-Teaching-and-Learning20191012-85439-1fft6xx.pdf [Accessed 11th
January 2023].
Agile Manifesto. (2001). Manifesto for Agile Software Development. Agilemanifesto.org. Available
from https://agilemanifesto.org/ [Accessed 11th January 2023].
Ainscow, M. (2007). Taking an inclusive turn. Journal of Research in Special Educational Needs, 7
(1), 3–7. Available from https://doi.org/10.1111/j.1471-3802.2007.00075.x [Accessed 9 January
2023].
Alberta Education (2017). Indicators of inclusive schools : Continuing the conversation. Indicators
of inclusive schools : continuing the conversation - Open Government. Available from
https://open.alberta.ca/publications/9781460107157 [Accessed 10 January 2023].
Alquraini, T.A. and Rotatori, A.F. (2014). Special Education International Perspectives: Practices
Across The Globe. Special Education International Perspectives: Practices Across the Globe,ii.
Available from https://doi.org/10.1108/S0270-401320140000028005 [Accessed 9 January 2023].
Amor, A.M. et al. (2018). International Perspectives and Trends in research on Inclusive Education:
A systematic review. International Journal of Inclusive Education, 23 (12), 1277–1295. Available
from https://doi.org/10.1080/13603116.2018.1445304 [Accessed 9 January 2023].
Anoyiannakis, K. (2013). Using technology to support individuals with ASD: A review of the
literature. ScholarWorks@GVSU. Available from
SEMICOLON - SE - 62 |I
References
https://scholarworks.gvsu.edu/honorsprojects/203/?utm_source=scholarworks.gvsu.edu%2Fhonorsp
rojects%2F203&utm_medium=PDF&utm_campaign=PDFCoverPages [Accessed 10 January 2023].
APA (1980). APA (1980) Diagnostic and statistical manual of mental disorders (3rd edition) (DSM-
III). American Psychiatric Association, Washington DC. - references - scientific research publishing.
Available from
https://www.scirp.org/(S(351jmbntvnsjt1aadkposzje))/reference/ReferencesPapers.aspx?ReferenceI
D=947108 [Accessed 10 January 2023].
APA (1994). Apa PsycNet. American Psychological Association. American Psychological

Association. Available from https://psycnet.apa.org/record/1994-97698-000 [Accessed 10 January
2023].
APA (2013). Apa PsycNet. American Psychological Association. American Psychological

Association. Available from https://psycnet.apa.org/record/2013-14907-000 [Accessed 10 January
2023].
Barned, N.E., Knapp, N.F. and Neuharth-Pritchett, S. (2011). Knowledge and attitudes of early
childhood preservice teachers regarding the inclusion of children with autism spectrum disorder.
Journal of Early Childhood Teacher Education, 32 (4), 302–321. Available from
https://doi.org/10.1080/10901027.2011.622235 [Accessed 9 January 2023].
Barned, N.E., Knapp, N.F. and Neuharth-Pritchett, S. (2011). Knowledge and attitudes of early
childhood preservice teachers regarding the inclusion of children with autism spectrum disorder.
Journal of Early Childhood Teacher Education, 32 (4), 302–321. Available from
https://doi.org/10.1080/10901027.2011.622235 [Accessed 9 January 2023].
Bentley-Williams, R. et al. (2016). Collaborative Partnership: Developing pre-service teachers as

inclusive practitioners to support students with disabilities. International Journal of Disability,
Development and Education, 64 (3), 270–282. Available from
https://doi.org/10.1080/1034912X.2016.1199851 [Accessed 9 January 2023].
Berkovits, L., Eisenhower, A. and Blacher, J. (2016). Emotion regulation in young children with
autism spectrum disorders. Journal of Autism and Developmental Disorders, 47 (1), 68–79. Available
from https://doi.org/10.1007/s10803-016-2922-2 [Accessed 9 January 2023].
Boehm, B., Turner, R., & Chulani, S. (2018). Lean software development: An agile toolkit. Addison-
Wesley Professional.
SEMICOLON - SE - 62 | II
References
Bondy, A. and Frost, L. (2001). The Picture Exchange Communication System. Behavior
Modification, 25 (5), 725–744. Available from https://doi.org/10.1177/0145445501255004
[Accessed 9 January 2023].
Bowe, F. (1970). Making inclusion work. Semantic Scholar. Available from

https://www.semanticscholar.org/paper/Making-Inclusion-Work-
Bowe/384ea3df3707ef2e50a6f588d63d0ab705fd9eb5 [Accessed 10 January 2023].
Brede, J. et al. (2017). Excluded from school: Autistic students’ experiences of school exclusion and
subsequent re-integration into school. Autism & Developmental Language Impairments, 2
239694151773751. Available from https://doi.org/10.1177/2396941517737511 [Accessed 9 January
2023].
Brownell, M.T. et al. (2005). Critical features of special education teacher preparation. The Journal
of Special Education, 38 (4), 242–252. Available from
https://doi.org/10.1177/00224669050380040601 [Accessed 9 January 2023].
Brue, A., Wilmshurst, L. and Person (2018). The Complete Guide to Special Education: Expert
Advice on evaluations. Taylor & Francis. Taylor & Francis. Available from
https://www.taylorfrancis.com/books/mono/10.4324/9781315111568/complete-guide-special-
education-linda-wilmshurst-alan-brue [Accessed 10 January 2023].
Brunsting, N.C., Sreckovic, M.A. and Lane, K.L. (2014). Special education teacher Burnout: A
synthesis of research from 1979 to 2013. Education and Treatment of Children, 37 (4), 681–711.
Available from https://doi.org/10.1353/etc.2014.0032 [Accessed 9 January 2023].
Busby, R. et al. (2011). Teaching elementary children with autism: Addressing teacher challenges
and preparation needs. Rural Educator. National Rural Education Association. Web site:
http://www.nrea.net. Available from https://eric.ed.gov/?id=EJ987618 [Accessed 10 January 2023].
Busby, R. et al. (2011). Teaching elementary children with autism: Addressing teacher challenges
and preparation needs. Rural Educator. National Rural Education Association. Web site:
http://www.nrea.net. Available from https://eric.ed.gov/?id=EJ987618 [Accessed 10 January 2023].
Busby, R. et al. (2018). Teaching elementary children with autism. The Rural Educator, 33 (2), .
Available from https://doi.org/10.35608/ruraled.v33i2.416 [Accessed 9 January 2023].
SEMICOLON - SE - 62 | III
References
Canada, P.H.A.of (2018). Government of Canada. Canada.ca. / Gouvernement du Canada. Available

from https://www.canada.ca/en/public-health/services/publications/diseases-conditions/autism-
spectrum-disorder-children-youth-canada-2018.html [Accessed 10 January 2023].
Carr, E.G. et al. (2002). Positive behavior support. Journal of Positive Behavior Interventions, 4 (1),
4–16. Available from https://doi.org/10.1177/109830070200400102 [Accessed 9 January 2023].
Chandler-Olcott, K. and Kluth, P. (2009). Why everyone benefits from including students with autism
in Literacy Classrooms. The Reading Teacher, 62 (7), 548–557. Available from
https://doi.org/10.1598/RT.62.7.1 [Accessed 9 January 2023].
Coman, D. et al. (2012). Commitment to classroom model philosophy and burnout symptoms among
high fidelity teachers implementing preschool programs for children with autism spectrum disorders.
Journal of Autism and Developmental Disorders, 43 (2), 345–360. Available from
https://doi.org/10.1007/s10803-012-1573-1 [Accessed 9 January 2023].
Conderman, G. and Johnston-Rodriguez, S. (2009). Beginning teachers' views of their collaborative

roles. Preventing School Failure: Alternative Education for Children and Youth, 53 (4), 235–244.
Available from https://doi.org/10.3200/psfl.53.4.235-244 [Accessed 9 January 2023].
Conderman, G. and Johnston-Rodriguez, S. (2009). Beginning teachers' views of their collaborative

roles. Preventing School Failure: Alternative Education for Children and Youth, 53 (4), 235–244.
Available from https://doi.org/10.3200/psfl.53.4.235-244 [Accessed 9 January 2023].
CoughDrop (2023). Every voice should be heard. CoughDrop. Available from

https://www.mycoughdrop.com/ [Accessed 10 January 2023].
Creswell, J.W. and Creswell, J.D. (2018). Research design: Qualitative, quantitative, and Mixed
Methods Approaches, 5th ed. Thousand Oaks, California: SAGE Publications. Available from
http://library.lol/main/408D5ADBE64C070B0A17BB3DB36B5A47 [Accessed 12th January]
Da Fonte, M.A. and Barton-Arwood, S.M. (2017). Collaboration of general and special education
teachers: Perspectives and strategies. Intervention in School and Clinic, 53 (2), 99–106. Available
from https://doi.org/10.1177/1053451217693370 [Accessed 9 January 2023].
Davis, J. (2016). Inclusion of students with high need autism: General education teachers' attitudes
and beliefs versus practice. ScholarWorks@Arcadia. Available from
https://scholarworks.arcadia.edu/grad_etd/13/ [Accessed 10 January 2023].
SEMICOLON - SE - 62 | IV
References
DeFilippis, M. (2018). Depression in children and adolescents with autism spectrum disorder.
Children, 5 (9), 112. Available from https://doi.org/10.3390/children5090112 [Accessed 9 January
2023].
Devlin, J. et al. (2019). Bert: Pre-training of deep bidirectional Transformers for language
understanding. arXiv.org. Available from https://arxiv.org/abs/1810.04805 [Accessed 10 January
2023].
Dowling, N. et al. (2020). Prevalence of autism spectrum disorder among children aged 8 years -
autism and Developmental Disabilities Monitoring Network, 11 sites, United States, 2016. Centers
for Disease Control and Prevention. Centers for Disease Control and Prevention. Available from
https://www.cdc.gov/mmwr/volumes/69/ss/ss6904a1.htm [Accessed 10 January 2023].
Eldar, E., Talmor, R. and Wolf‐Zukerman, T. (2009). Successes and difficulties in the individual
inclusion of children with autism spectrum disorder (ASD) in the eyes of their coordinators.
International Journal of Inclusive Education, 14 (1), 97–114. Available from
Emam, M.M. and Farrell, P. (2009). Tensions experienced by teachers and their views of support for
pupils with Autism Spectrum Disorders in mainstream schools. European Journal of Special Needs
Education, 24 (4), 407–422. Available from https://doi.org/10.1080/08856250903223070 [Accessed
9 January 2023].
Erpelding, J.M. (2011). Visual teaching strategies for autistic students in inclusive general education
classrooms. ProQuest LLC. ProQuest LLC. 789 East Eisenhower Parkway, P.O. Box 1346, Ann
Arbor, MI 48106. Tel: 800-521-0600; Web site: http://www.proquest.com/en-
US/products/dissertations/individuals.shtml. Available from https://eric.ed.gov/?id=ED551304
Finke, E.H. et al. (2009). “all children can and should have the opportunity to learn”: General
Education Teachers' perspectives on including children with autism spectrum disorder who require
AAC. Augmentative and Alternative Communication, 25 (2), 110–122. Available from
Flower, A., McKenna, J.W. and Haring, C.D. (2016). Behavior and classroom management: Are
teacher preparation programs really preparing our teachers? Preventing School Failure: Alternative
SEMICOLON - SE - 62 |V
References
Education for Children and Youth, 61 (2), 163–169. Available from

https://doi.org/10.1080/1045988X.2016.1231109 [Accessed 9 January 2023].
Fowler, M. (2015). UML distilled : a brief guide to the standard object modeling language. Boston:
Addison-Wesley. Available from https://cloudflare-
ipfs.com/ipfs/bafykbzacebpuaijzusudrzbnl5fjw7oguby533hoymt6lw5pi5vhnw2lpwcji?filename=M
artin%20Fowler%20-
%20UML%20Distilled_%20A%20Brief%20Guide%20to%20the%20Standard%20Object%20Mod
eling%20Language-Addison-Wesley%20Professional%20%282003%29.pdf [Accessed 12th
January].
Frederickson, N., Jones, A.P. and Lang, J. (2010). Inclusive provision options for pupils on the autistic
spectrum. Journal of Research in Special Educational Needs, 10 (2), 63–73. Available from
https://doi.org/10.1111/j.1471-3802.2010.01145.x [Accessed 9 January 2023].
Frontline Education (2022). School administration software for K-12. Frontline Education. Available
from https://www.frontlineeducation.com/ [Accessed 10 January 2023].
Ghanouni, P. et al. (2019). The use of technologies among individuals with autism spectrum
disorders: Barriers and challenges. Journal of Special Education Technology, 35 (4), 286–294.
Available from https://doi.org/10.1177/0162643419888765 [Accessed 9 January 2023].
Google Cloud, (2023). Google cloud documentation | documentation. Google. Available from
https://cloud.google.com/docs [Accessed 10 January 2023].
Graham, L.J. and Sweller, N. (2011). The Inclusion Lottery: Who's in and who's out? tracking
inclusion and exclusion in New South Wales Government Schools. International Journal of Inclusive
Education, 15 (9), 941–953. Available from https://doi.org/10.1080/13603110903470046 [Accessed
9 January 2023].
Gray, C.A. and Garand, J.D. (1993). Social Stories: Improving responses of students with autism with
accurate social information. Focus on Autistic Behavior, 8 (1), 1–10. Available from
Guldberg, K. (2010). Educating children on the autism spectrum: Preconditions for inclusion and
notions of ‘best autism practice’ in the early years. British Journal of Special Education, 37 (4), 168–
174. Available from https://doi.org/10.1111/j.1467-8578.2010.00482.x [Accessed 9 January 2023].
SEMICOLON - SE - 62 | VI
References
Gómez-Marí, I., Sanz-Cervera, P. and Tárraga-Mínguez, R. (2022). Teachers’ attitudes toward autism
spectrum disorder: A systematic review. Education Sciences, 12 (2), 138. Available from
https://doi.org/10.3390/educsci12020138 [Accessed 9 January 2023].
Hettiaarachi, S. et al. (2017). Including all? perceptions of mainstream teachers on inclusive education
in the Western Province of Sri Lanka. International Journal of Special Education. International
Journal of Special Education. 2889 Highbury Street, Vancouver, BC V6R 3T7, Canada. Web site:
http://www.internationalsped.com/. Available from https://eric.ed.gov/?id=EJ1185612 [Accessed 10
January 2023].
Higginson, R. and Chatfield, M. (2011). Together we can do it: A professional development project
for regular teachers' of children with autism spectrum disorder. Kairaranga. Massey University.
Private Bag 11 222; Palmerston North 4442, New Zealand. Tel: +64-6-351-3396; Fax: +64-6-351-
3472; email: [email protected]; Web site:
http://www.massey.ac.nz/massey/learning/departments/school-curriculum-
pedagogy/kairaranga/kairaranga_home.cfm. Available from https://eric.ed.gov/?id=EJ994983
Hughes, H., Combes, B.H. and Metha, S.S. (2012). Managing autism: Knowledge and training in
autism spectrum disorders among special education administrators in Texas. Journal of Special
Education Leadership. Council of Administrators of Special Education. Fort Valley State University,
1005 State University Drive, Fort Valley, GA 31030. Tel: 478-825-7667; Fax: 478-825-7811; Web
site: http://www.casecec.org. Available from https://eric.ed.gov/?id=EJ997650 [Accessed 10 January
2023].
Hyman, S.L. et al. (2020). Identification, evaluation, and management of children with autism
spectrum disorder. Pediatrics, 145 (1), . Available from https://doi.org/10.1542/peds.2019-3447
Kaushik, V. and Walsh, C.A. (2019). Pragmatism as a Research Paradigm and Its Implications for
Social Work Research. Social Sciences, 8 (9), 255. Available from
https://doi.org/10.3390/socsci8090255 [Accessed 12th January 2023].
King, M. and Bearman, P. (2009). Diagnostic change and the increased prevalence of autism.
International Journal of Epidemiology, 38 (5), 1224–1234. Available from
https://doi.org/10.1093/ije/dyp261 [Accessed 9 January 2023].
SEMICOLON - SE - 62 | VII
References
Kurth, J. and Mastergeorge, A.M. (2010). Impact of setting and instructional context for adolescents
with autism. The Journal of Special Education, 46 (1), 36–48. Available from
Lindsay, S. et al. (2013). Exploring teachers' strategies for including children with autism spectrum
disorder in mainstream classrooms. International Journal of Inclusive Education, 18 (2), 101–122.
Available from https://doi.org/10.1080/13603116.2012.758320 [Accessed 9 January 2023].
Lindsay, S. et al. (2013). Exploring teachers' strategies for including children with autism spectrum
disorder in mainstream classrooms. International Journal of Inclusive Education. Routledge.
Available from: Taylor & Francis, Ltd. 325 Chestnut Street Suite 800, Philadelphia, PA 19106. Tel:
800-354-1420; Fax: 215-625-2940; Web site: http://www.tandf.co.uk/journals. Available from
https://eric.ed.gov/?id=EJ1029923#:~:text=Teachers%20had%20several%20recommendations%20f
or,5)%20building%20a%20climate%20of [Accessed 10 January 2023].
Liu, Y. et al. (2019). Roberta: A robustly optimized Bert pretraining approach. arXiv.org. Available
from https://arxiv.org/abs/1907.11692v1 [Accessed 10 January 2023].
Loiacono, V. and Valenti, V. (2009). General education teachers need to be prepared to co-teach the
increasing number of children with autism in inclusive settings. International Journal of Special
Education. International Journal of Special Education. 2889 Highbury Street, Vancouver, BC V6R
3T7, Canada. Web site: http://www.internationaljournalofspecialeducation.com Available from
https://eric.ed.gov/?id=EJ909033 [Accessed 10 January 2023].
Lovaas, O.I. (1987). Behavioral treatment and normal educational and intellectual functioning in
young autistic children. Journal of Consulting and Clinical Psychology, 55 (1), 3–9. Available from
https://doi.org/10.1037/0022-006X.55.1.3 [Accessed 9 January 2023].
Mandell, D.S. et al. (2005). The prevalence and correlates of abuse among children with autism served
in comprehensive community-based mental health settings. Child Abuse & Neglect, 29 (12), 1359–
1372. Available from https://doi.org/10.1016/j.chiabu.2005.06.006 [Accessed 9 January 2023].
Marks, S.U. et al. (2003). Instructional management tips for teachers of students with autism spectrum
disorder (ASD). TEACHING Exceptional Children, 35 (4), 50–54. Available from
McCray, E.D. and McHatton, P.A. (2010). "Less afraid to have "them" in my classroom":
Understanding pre-service general educators' perceptions about inclusion. Teacher Education
SEMICOLON - SE - 62 | VIII
References
Quarterly. Caddo Gap Press. 3145 Geary Boulevard PMB 275, San Francisco, CA 94118. Tel: 415-
666-3012; Fax: 415-666-3552; e-mail: [email protected]; Web site: http://www.caddogap.com.
Available from https://eric.ed.gov/?id=EJ960622 [Accessed 10 January 2023].
McCrimmon, A.W. (2014). Inclusive Education in Canada. Intervention in School and Clinic, 50 (4),
234–237. Available from https://doi.org/10.1177/1053451214546402 [Accessed 9 January 2023].
McKeon, B., Alpern, C.S. and Zager, D. (2012). Promoting academic engagement for college students
with autism spectrum disorder. Journal of Postsecondary Education and Disability. Association on
Higher Education and Disability. 107 Commerce Center Drive Suite 204, Huntersville, NC 28078.
Tel: 704-947-7779; Fax: 704-948-7779; e-mail: [email protected]; Web site:
http://www.ahead.org/publications/jped. Available from https://eric.ed.gov/?id=EJ1026894
McLeskey, J. et al. (2010). Are we moving toward educating students with disabilities in less
restrictive settings? The Journal of Special Education, 46 (3), 131–140. Available from
McLeskey, J., Waldron, N.L. and Redd, L. (2012). A case study of a highly effective, inclusive
elementary school. The Journal of Special Education, 48 (1), 59–70. Available from
McNicholl, A. et al. (2019). The impact of assistive technology use for students with disabilities in
higher education: A systematic review. Disability and Rehabilitation: Assistive Technology, 16 (2),
130–143. Available from http://dx.doi.org/10.1080/17483107.2019.1642395 [Accessed 9 January
2023].
Merriman, D. et al. (2016). The effects of group coaching on the homework problems experienced
by secondary students with and without disabilities. Psychology in the Schools, 53 (5), 457–470.
Available from https://onlinelibrary.wiley.com/doi/10.1002/pits.21918 [Accessed 9 January 2023].
Michael Quinn Patton. (2011). Developmental evaluation : applying complexity concepts to enhance
innovation and use. New York: Guilford. Available from https://cloudflare-
ipfs.com/ipfs/bafykbzacedu3ktlqrjbfubf5ncopz6atgh3akowdfjtfg7gfsy6ic4eroojwy?filename=Mich
ael%20Quinn%20Patton%20-
%20Developmental%20Evaluation_%20Applying%20Complexity%20Concepts%20to%20Enhanc
SEMICOLON - SE - 62 | IX
References
e%20Innovation%20and%20Use-Guilford%20Press%20%282010%29.pdf [Accessed 12th January

2023].
National Center for Education Statistics (2020). . Special education management software: A review
of current offerings and best practices. . Special education management software: A review of current
offerings and best practices. . Available from
https://nces.ed.gov/pubsearch/pubsinfo.asp?pubid=202150 [Accessed 10 January 2023].
National Institute on Deafness and Other Communication Disorders (2021). Assistive devices for
people with hearing, voice, speech, or language disorders. National Institute of Deafness and Other
Communication Disorders. U.S. Department of Health and Human Services. Available from
https://www.nidcd.nih.gov/health/assistive-devices-people-hearing-voice-speech-or-language-
disorders [Accessed 10 January 2023].
Nuske, H.J. et al. (2019). Do student characteristics affect teachers’ decisions to use 1:1 instruction?
Journal of Autism and Developmental Disorders, 49 (7), 2864–2872. Available from
https://doi.org/10.1007/s10803-019-04004-1 [Accessed 9 January 2023].
Olszewski, A., Smith, E. and Franklin, A.D. (2022). Speech-language pathologists' feelings and
practices regarding technological apps in School Service Delivery. Language, Speech, and Hearing
Services in Schools, 53 (4), 1051–1073. Available from https://doi.org/10.1044/2022_LSHSS-21-
00150 [Accessed 9 January 2023].
OpenAI (2022). Dall·E: Creating images from text. OpenAI. OpenAI. Available from
https://openai.com/blog/dall-e/ [Accessed 9 January 2023].
Pagaling, R.J. (2021). Exploring resources provided to general education teachers to enhance their
understanding and capacity when working with students with autism spectrum disorder (ASD).
Digital Object Identifier System. Werklund School of Education. Available from
https://doi.org/10.11575/PRISM/38532 [Accessed 10 January 2023].
Pancsofar, N. and Petroff, J.G. (2013). Professional development experiences in co-teaching. Teacher
Education and Special Education: The Journal of the Teacher Education Division of the Council for
Exceptional Children, 36 (2), 83–96. Available from https://doi.org/10.1177/0888406412474996
[Accessed 9 December 2023].
Parsons, L.D., Miller, H. and Deris, A.R. (2015). The effects of special education training on educator
efficacy in classroom management and inclusive strategy use for students with autism in inclusion
SEMICOLON - SE - 62 |X
References
classes. Journal of the American Academy of Special Education Professionals. American Academy
of Special Education Professionals. 3642 East Sunnydale Drive, Chandler Heights, AZ 85142. Tel:
800-754-4421; Fax: 800-424-0371; e-mail: [email protected]; Web site: http://www.aasep.org.
Available from https://eric.ed.gov/?id=EJ1129741 [Accessed 10 January 2023].
Pulkkinen, J. and Jahnukainen, M. (2015). Finnish reform of the funding and provision of special
education: The Views of principals and Municipal Education Administrators. Educational Review,
68 (2), 171–188. Available from https://doi.org/10.1080/00131911.2015.1060586 [Accessed 9
January 2023].
Radford, A. (2021). Better language models and their implications. OpenAI. OpenAI. Available from
https://openai.com/blog/better-language-models/ [Accessed 10 January 2023].
Ramesh, A. et al. (2022). Hierarchical text-conditional image generation with clip latents. arXiv.org.
Available from https://doi.org/10.48550/ARXIV.2204.06125 [Accessed 10 January 2023].
Ross-Hill, R. (2009). Teacher attitude towards inclusion practices and special needs students. Journal
of Research in Special Educational Needs, 9 (3), 188–198. Available from
Ruel, M.-P., Poirier, N. and Japel, C. (2017). La perception d’enseignantes du primaire quant à
l’intégration en classe ordinaire d’élèves présentant un trouble du spectre de l’autisme (TSA). Revue
de psychoéducation, 44 (1), 37–61. Available from https://doi.org/10.7202/1039270ar [Accessed 9
January 2023].
Ryan, G. (2018). Introduction to positivism, interpretivism and critical theory. Nurse Researcher, 25
(4), 41–49. Available from https://doi.org/10.7748/nr.2018.e1466 [Accessed 12th January 2023].
Sailor, W. (2014). Advances in schoolwide inclusive school reform. Remedial and Special Education,
36 (2), 94–99. Available from https://doi.org/10.1177/0741932514555021 [Accessed 9 January
2023].
Sansosti, F.J., Powell-Smith, K.A. and Kincaid, D. (2004). A research synthesis of social story
interventions for children with autism spectrum disorders. Focus on Autism and Other Developmental
Disabilities, 19 (4), 194–204. Available from https://doi.org/10.1177/10883576040190040101
SEMICOLON - SE - 62 | XI
References
Schaefer, G. (2016). Clinical genetic aspects of autism spectrum disorders. International Journal of
Molecular Sciences, 17 (2), 180. Available from https://doi.org/10.3390/ijms17020180 [Accessed 9
January 2023].
Scruggs, T.E. and Mastropieri, M.A. (1996). Teacher perceptions of mainstreaming/inclusion, 1958–
1995: A research synthesis. Exceptional Children, 63 (1), 59–74. Available from
SEN Educator Recruiter. (2022). Interviewee. [Interview]. Online video. Available at:
https://drive.google.com/drive/folders/1fWyMpOakdKiOjdP-2YF6lWfNr3jetM5v?usp=share_link
SEN Teacher. (2022). Interviewee. [Interview]. Online video. Available at:

https://drive.google.com/drive/folders/1l7wfj2M9cMR6AqYO5i1owEw6M6HdABf1?usp=share_li
nk
Slee, R. (1998). The Politics of Theorizing Special Education. the UWA Profiles and Research
Repository. Routledge. Available from https://research-repository.uwa.edu.au/en/publications/the-
politics-of-theorizing-special-education [Accessed 10 January 2023].
Strang, J.F. et al. (2012). Depression and anxiety symptoms in children and adolescents with autism
spectrum disorders without intellectual disability. Research in Autism Spectrum Disorders, 6 (1),
406–412. Available from https://doi.org/10.1016/j.rasd.2011.06.015 [Accessed 9 January 2023].
Sullivan, T.N. and Bradshaw, C.P. (2012). Introduction to the special issue of behavioral disorders:
Serving the needs of youth with disabilities through school-based violence prevention efforts.
Behavioral Disorders, 37 (3), 129–132. Available from
Symes, W. and Humphrey, N. (2011). The deployment, training and teacher relationships of teaching
assistants supporting pupils with Autistic Spectrum Disorders (ASD) in Mainstream Secondary
Schools. British Journal of Special Education, 38 (2), 57–64. Available from
Syriopoulou-Delli, C.K. et al. (2011). Teachers’ perceptions regarding the management of children
with autism spectrum disorders. Journal of Autism and Developmental Disorders, 42 (5), 755–768.
Available from https://doi.org/10.1007/s10803-011-1309-7 [Accessed 9 January 2023].
SEMICOLON - SE - 62 | XII
References
Tissot, C. and Evans, R. (2003). Visual teaching strategies for children with autism. Early Child
Development and Care, 173 (4), 425–433. Available from
Toran, H. et al. (1970). [PDF] the preparation, knowledge and self reported competency of Special
Education Teachers Regarding students with autism: Semantic scholar. Pertanika journal of social
science and humanities. Available from https://www.semanticscholar.org/paper/The-
preparation%2C-knowledge-and-self-reported-of-Toran-
Westover/1af685316e096b80b4e888ea466fac749839156a [Accessed 10 January 2023].
Trainee SEN Teacher. (2022). Interviewee. [Interview]. Online video. Available at:
https://drive.google.com/drive/folders/14dRQI0okn9l4FopOBtkrKBQ5u4rJjg7k?usp=share_link
Unesco and UNESCO Office in Beirut UNESCO (1970). Inclusion in education. UNESCO.org.
Available from https://www.unesco.org/en/education/inclusion [Accessed 10 January 2023].
University of Kansas (2020). Assistive Technology for Kansans. Available from https://atk.ku.edu/
VanBergeijk, E., Klin, A. and Volkmar, F. (2008). Supporting more able students on the autism
spectrum: College and beyond. Journal of Autism and Developmental Disorders, 38 (7), 1359–1370.
Available from https://doi.org/10.1007/s10803-007-0524-8 [Accessed 9 January 2023].
W Lawrence Neuman. (2014). Social research methods : qualitative and quantitative approaches, 7th
ed. Boston, Ma: Pearson Education, Inc. Available from https://www.pdfdrive.com/social-research-
methods-qualitative-and-quantitative-approaches-e55615757.html [Accessed 12th January 2023].
Warnock, M. (2008). Special educational needs: A new look. Challenges for Inclusion,43–65.
Available from https://doi.org/10.1163/9789087903039_004 [Accessed 9 January 2023].
Whalon, K. and Hart, J.E. (2010). Adapting an evidence-based reading comprehension strategy for
learners with autism spectrum disorder. Intervention in School and Clinic, 46 (4), 195–203. Available
from https://doi.org/10.1177/1053451210389036 [Accessed 9 January 2023].
WhatsApp Business API (no date). Twilio. Available from https://www.twilio.com/whatsapp

SEMICOLON - SE - 62 | XIII
References
Yourdon, E. and Constantine, L.L. (2016). Structured design: Fundamentals of a discipline of

computer program and system design. Pearson Education. Available from
https://www.pdfdrive.com/structureddesign_edw-e35089460.html [Accessed 12th January 2023].
Zhang, C. and Hu, B. (2015). Inclusion as an approach and process for promoting acceptance and
success: Comparative perspectives between the United States and China. International Journal of
Early Childhood Special Education, 7 (2), 238. Available from
https://doi.org/10.20489/intjecse.83870 [Accessed 9 January 2023].
Zhang, L. et al. (2020). An interactive multi-task learning framework for next poi recommendation
with uncertain check-ins. IJCAI. Available from https://www.ijcai.org/Proceedings/2020/491
[Accessed 9 January 2023]
SEMICOLON - SE - 62 | XIV
Appendix
APPENDIX
Appendix - Section A - Literature Review
Appendix Section A.1 - Concept Graph
Figure 28 :Concept Graph(Self Composed)
SEMICOLON - SE - 62 | XV
Appendix
Appendix - Section B - Methodologies
Appendix Section B.1 - Work breakdown Structure
Figure 8 :Teamwork breakdown structure(Self Composed)
SEMICOLON - SE - 62 | XVI
Appendix
Appendix Section B.2 - Gantt Chart
Figure 9 : Gantt Chart(Self Composed)
SEMICOLON - SE - 62 | XVII
Appendix
Appendix Section B.3 - Project Management Methodology
Appendix Section B.3.1 - ClickUP
Figure 29 :Section breakdowns in ClickUp
Figure 30 :Task Checklist in ClickUp
SEMICOLON - SE - 62 | XVIII
Appendix
Appendix Section B.3.2 - Slack
Figure 31 :Collaboration between G-Drive and Slack
Appendix Section B.3.3 - Google Calander - Scheduled Google Meet meetings
Figure 32 :Snip of allocations for the weekly meeting schedule
Appendix Section B.3.4 - Table of Meeting minutes
SEMICOLON - SE - 62 | XIX
Appendix
Month Date Time
October Monday 3rd, 10th, 17th, 24th 8:30am – 12:00am

(Physical Meeting)
Tuesday 4th, 11th, 18th, 25th 2:30am – 4:30pm
Thursday 13th, 20th 7:00pm – 9:30pm
Sunday 2nd, 9th, 16th 7:00pm – 9:00pm
23rd, 30th 8:30pm – 10:00pm
November Monday 7th, 14th 7:30pm – 9:30pm

21st, 28th 7:00pm – 9:00pm
Tuesday 8th, 15th, 22th 10:30am – 12:30pm
Wednesday 16th 8:00pm – 9:30pm
30th 1:00pm – 3:00pm
23rd 7:30pm – 9:30pm
Saturday 5th 6:30pm – 9:30pm
12th, 19th 8:00pm – 9:30pm
December Monday 5th, 12th 1:00pm – 4:00pm
19th 7:30pm – 9:30pm
Thursday 8th, 15th 5:30pm – 7:30pm
22th, 29th 10:30am – 12:30pm
SEMICOLON - SE - 62 | XX
Appendix
Tuesday 13th, 20th 7:30pm – 10:30pm
Saturday 10th, 24th 8:00pm – 11:30pm
Table 37 : Schedule for weekly meetings and additional sessions
SEMICOLON - SE - 62 | XXI
Appendix
Appendix - Section C - SRS
Appendix - Section C.1 - Questionnaire
SEMICOLON - SE - 62 | XXII
Appendix
SEMICOLON - SE - 62 | XXIII
Appendix
Figure 33 :Snip of the survey
Appendix - Section C.2 - Questionnaire Findings
Question Select your age group
Aim of the question to determine the age group the survey's

participants belong to as well as the age
range of those who are aware of autism.
Observation
SEMICOLON - SE - 62 | XXIV
Appendix
The survey's entire sample ranges in age

from 21 to 40. There have been no
participants who are younger than 18 or
older than 40.
Conclusion The survey's entire sample ranges in age

from 21 to 40. People under the age of 18
and those over 40 have not responded to
the survey, maybe because they lack IT
expertise.
Question Do you have an idea about Autism?
Aim of the question To know how many people at least have an

idea about Autism.
Observation
It was observed that all participants have an

idea about what Autism is.
Conclusion All of the participants of the survey have an

idea about Autism.
Question Have you had firsthand experience with an

Autistic child?
Aim of the Question To know how many participants have had a

real time experience with an Autistic child.
Observation
SEMICOLON - SE - 62 | XXV
Appendix
It was observed that none of the people has

had a firsthand experience with an Autistic
child.
Conclusion It should be understood that none of the

poll respondents have any knowledge of an
autistic child's behaviour because none of
them have first-hand experience with one.
Question What do you think of creating a common

application for teachers,Autistic students
and caregivers?
Aim of the question To learn what the general public thinks

about the system's value for the three
involved parties—teachers, autistic pupils,
and caretakers.
Observation
All the participants of the survey think that

SupAut will be a very useful application for
teachers,Autistic students and caregivers.
Conclusion From the responses it could be proven that

the peoples opinion is that SupAut will be a
useful application for the Autistic
students,teachers and caregivers.
Question Do you think that an online application like

SupAut will help the middle school teacher
SEMICOLON - SE - 62 | XXVI
Appendix
to tutor Autistic children?
Aim of the question To know about what people think of

SupAut from the main users(middle school
teachers)side.
Observation
It was found that, overall, 55.6% of

participants believe that SupAut will make
tutoring for middle school teachers simple,
compared to 44.6% who are unsure about
the application's ability to do so. However,
nobody believes that it won't be useful
while instructing children with autism.
Conclusion The overall results show that individuals

believe SupAut will be highly helpful for
middle school teachers while tutoring the
autistic students.
Question Do you think this online application will

facilitate the Autistic kids to learn their
subject material in a more effective way?
Aim of the question To know about what people think of

SupAut from the Autistic students side.
Observation
33.3% of respondents believe that SupAut

will be a helpful online tool for helping
autistic children study their subject content
effectively. 66.7% of respondents are
unsure about whether SupAut will be an
effective online tool for helping autistic
children study their subject.
SEMICOLON - SE - 62 | XXVII
Appendix
Conclusion The responses indicate that the majority of

individuals believe SupAut will be a
valuable online tool for helping autistic
children understand their subject content.
Question What are the most important features that

should be included in SupAut?
Aim of the question To get a clear idea about what features

should be given priority when developing
the system.
Observation
77.8% of respondents believe that all

features should be prioritized equally.
Lesson planning, resource listing, and the
creation of a shared platform for teachers,
students, and caregivers, according to
22.2% of respondents, should be given
higher emphasis. Transitioning from
unanticipated changes, according to 11.1%,
should be given higher attention.
Conclusion All the features should be considered as

high priority and should be included in the
SupAut system.
Question What do you think of including the VR

technology in SupAut?
Aim of the question To get a response from people about

including the VR technology in SupAut?
Observation
SEMICOLON - SE - 62 | XXVIII
Appendix
While 44.4% are unsure, 55.6% believe

that adding VR technology to SupAut is
very valuable. Nobody believes that
SupAut's integration of VR technology
won't be beneficial.
Conclusion People clearly want VR technology in

SupAut based on the replies, but the notion
is one that the developing team would
implement after creating the basic system.
Question What do you think of including chatbots in

SupAut?
Aim of the question To get the point of view from the people
about the inclusion of a chatbot in SupAut.
Observation
The installation of a chatbot for the SupAut

system is viewed as being extremely useful
by 66.7% of people. 33.3% of the
population is unclear on the subject. None
of the respondents believed that the
addition of a chatbot would be ineffective.
Conclusion From the responses it could be proven that

the inclusion of a chatbot is really
necessary for the SupAut system.
Table 22: Questionnaire Findings table
Appendix - Section C.3 - Interview Findings
C.3.1 Interview 1 with a SEN Teacher (2022)
Question How is it to deal with Autistic Children in an inclusive classroom?
SEMICOLON - SE - 62 | XXIX
Appendix
Aim of the Question Understanding the mentality of Autistic children During study sessions
Observation It is difficult for regular teachers to adapt themselves in order to stick to

the Autistic children and stay beside them.
Every Child grasps knowledge at a different pace.
Most Autistic children needs support
Conclusion Each and every Autistic child should have a Caretaker who should stay
beside them in order to help in their needs . Therefore it is expected to
assign at least one caretaker to assist one child in the project
Question What are the challenges the Autistic child faces inside the classroom?
Aim of the Question Understanding the situations and challenges faced by Autistic children
Observation Inability to Participate in co-curricular , extra - curricular activities.

Not preferred to lead or participate in special event
Regular teachers pushing the child away from such events
Conclusion Regardless of being an autistic child or a regular child, everyone should

be motivated to ask and do/learn things they like. So the built in chatbot
will help the child ask and learn.
Question What are the things You wish that would help the Autistic children
Aim of the Question Getting ideas for the project which could help the Autistic children
Observation Since most of the autistic children take time to understand things it is
tough for the teachers
Caretakers should take spend more time in explaining in case if the
child had not understood something
Autistic children should be given extra study materials in order to learn
something in the right way
Conclusion Autistic Children should be cared for Extra , They need extra materials
to learn and improve themselves . Thus, Staff will be providing reports
SEMICOLON - SE - 62 | XXX
Appendix
regarding the progress of child
C.3.2 Interview 2 with a Trainee SEN Teacher (2022)
Question How does the SEN Children’s teaching techniques change throughout the
Primary school and middle school?
Aim of the Question Understanding how teachers technically handle the Autistic children with
new techniques
Observation Visual, Auditory , Reading methods differ from normal teaching

Students are taught with unique techniques so that they would understand.
Conclusion Different techniques should be used for the understanding of students

accordingly. So it’s expected to have innovative techniques in the project
Question Were SEN students included with the mainstream students?
Aim of the Question Getting to know if the SEN children will be able to learn properly even if
enough support is not given
Observation Some have behaviours which needs help

Children who are able to control themselves will be with the mainstream
children for specific subjects, but all the SEN children will be taught the
main subjects under special staff
Conclusion SEN students are handled according to their behaviours and how
controllable they are. Analysis of the child's progression is among the
main objectives of the project.
Question How do you get to know the interest or skill of the SEN child?
Aim of the Question Knowing techniques which will help in finding interest and skills of
Autistic children
Observation Exams and test are conducted to assess the skills and interests of autistic
children
SEMICOLON - SE - 62 | XXXI
Appendix
Conclusion Interests and skills can be assessed by conducting tests and exams.
Therefore questions which will be uploaded in the portal will help assess
the child's skills and interests.
C.3.3 Interview 3 with a SEN Recruiter (2022)
Question What do you look for in teachers when you recruit them?
Aim of the Question Getting to know the requirements of the staff who engage with autistic
children
Observation Characteristics , skill sets and work ethics matter when recruiting staff
Conclusion Staff who are recruited should have great characteristics and skills to
manage and educate the children. This response make clear that the
staff included in the project should hold a great character and better
skills
Question How does the experience of staff affect the child’s education?
Aim of the Question Comparison between experienced and amateur staff in Special
Education Needs
Observation Amateur Staff would struggle in handling students meanwhile the

experienced ones would find a way to handle and guide the child
Conclusion Experienced staff are much helpful in the guiding process of the
autistic children . Therefore, experienced staff will be recruited to
handle the progression and analysis.
Question What are the technologies used to help in assisting the autistic children
in education?
Aim of the Question Getting to know about the technologies which could help in assisting
the autistic children
Observation Some institutions have different educational plans where they focus
SEMICOLON - SE - 62 | XXXII
Appendix
mainly in the interests and skills of the child
Conclusion Planning according to the child's interests will help the child a lot. This
observation ensures that using various techniques in helping the autistic
children is a must consider part of the project
Appendix - Section C.4 - Literature Review Findings
Findings Description
DALL-E Suitable AI Art Generator API endpoint to help visualize concepts
GPT-3 Suitable Transformer-based language model API endpoint to detect

difficulty in comprehending or understanding questions and act as a
teacher replacement
Twilio Suitable Conversational AI to carry out the flow of conversation through

WhatsApp
Google Cloud Suitable Natural Language Processing (NLP) technique to detect anti-
Natural Language social sentiment and correct the student with the right conversing
standards fitting social norm
Table 26: Selection/ findings of Literature Review
SEMICOLON - SE - 62 | XXXIII
Appendix
Appendix - Section C.4 - Use case descriptions
Use Case Name Authenticate
Use Case ID UC-001
Description This use case describes how a teacher can login to the system
Priority High
Primary Actor teacher
Supporting Actors None
Pre-Conditions The web application should be installed to your device and a stable
internet connection is needed.
Trigger The teacher request authenticate
1. Open the web 3. System validates the details

application entered.
2. Enter the login details. 4. The system will proceed towards
the dashboard page.
Enters invalid information System displays errors

● Incorrect username ● Invalid password
● Incorrect password ● Invalid username
Users leave any of the fields ● System displays a message
empty. “ You cannot leave any of the fields
empty ”
SEMICOLON - SE - 62 | XXXIV
Appendix
None None
Exclusions None
Post Conditions System will display the main dashboard page.
Table 27 :Use Case descriptions for “Authentication”
Use Case Name Post Questions
Description The student will post a question to the teacher.
Priority High
Primary Actor Student
Supporting Actors Teacher
Pre-Conditions The student should have received a question/questions from the teacher.
Trigger In his or her WhatsApp desktop or mobile app, the teacher will be able to
see the pupils' questions.
1. The student will post a 3. The system will show the

question/provide student's queries or responses
answers . that were posted to the middle
2. The inquiry and school teacher.
response can be seen
by the teacher on his
SEMICOLON - SE - 62 | XXXV
Appendix
or her WhatsApp
desktop or mobile app.
4. The student does not 5. The system will not display

answer the anything in the WhatsApp
questions/post a mobile app/desktop version of
question. the teacher.
None None
Exclusions Rephrase
Post Conditions The student's response or query will be visible to the teacher in his or her
WhatsApp desktop or mobile app. The dashboard will keep track of the
chat activities.
Table 32 :Use case descriptions for “post questions”
SEMICOLON - SE - 62 | XXXVI
Appendix
Appendix - Section D - Design
Figure 27 :The Process Activity diagram(Self Composed)
SEMICOLON - SE - 62 | XXXVII
Appendix
Appendix - Section E - Work Breakdown Matrix
Chapter Chapter Breakdown Contributors
Chapter 1 - Introduction Chapter Overview Zainab Fahim
Problem Background Zainab
Problem Statement Zainab
Research Gap Zainab
Research Questions Zainab
Research Aim Zainab
Project Scope Zainab
Rich Picture Diagram Zainab
Objectives Adheeb
Resource Requirements Adheeb
Chapter Summary Zainab
Chapter 2 - Literature Chapter Overview Zainab Fahim

Review
Concept Graph Zainab
Domain Zainab
Tools and Techniques Zainab
Existing Solutions Zainab
Chapter Summary Zainab
Chapter 3 - Methodology Chapter Overview Adheeb Azahim, Mewan

Manodhya
Research Methodology Adheeb
Development Methodology Adheeb
Design Methodology Adheeb
Evaluation Methodology Adheeb
Project Management Adheeb, Mewan

Methodology
Team Work Breakdown Adheeb, Mewan
SEMICOLON - SE - 62 | XXXVIII
Appendix
Structure
Gantt chart diagram Mewan
Usage of Project Management Mewan

and collaboration software in
the project
Risks and Mitigation Adheeb
Chapter Summary Adheeb, Mewan
Chapter 4 - System Chapter Overview Sandil Munasinghe

Requirements Specification Minoli Dayarathne
(SRS)
Stakeholders Analysis Minoli
Selection of Requirement Sandil, Zainab

Elicitation Techniques
Analysis of Gathered Data Sandil, Adheeb
Context Diagram Sandil
Use Case Diagram Sandil , Zainab , Adheeb ,

Minoli
Use Case Descriptions Sandil
Domain Model Minoli, Zainab
Functional Requirements Sandil
Non-Functional Requirement Sandil
Chapter Summary Sandil
Chapter 5 - Social, Legal, Chapter Overview Mewan Manodhya

Ethical and Professional
Issues (SLEP)
SLEP Issues and Mitigation Mewan
Chapter Summary Mewan
Chapter 6 - System Chapter Overview Sandil Munasinghe, Zainab

Architecture & Design Fahim
Minoli Dayarathne, Adheeb
Azahim
System Architecture Design Zainab
SEMICOLON - SE - 62 | XXXIX
Appendix
Class Diagram Zainab, Sandil
Sequence Diagrams Sandil
Process Flow Chart Sandil
UI Design Zainab, Adheeb, Minoli,

Mewan, Sandil
Chapter Summary Sandil
Table 38 : Work Breakdown Matix table
SEMICOLON - SE - 62 | XL

Zainab's Report SRS

Uploaded by

Copyright:

Available Formats

Zainab's Report SRS

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Zainab's Report SRS

Uploaded by

Copyright:

Available Formats

INFORMATICS INSTITUTE OF TECHNOLOGY

Project SupAut: Teaching Using Effective Assistive Technology To Support

Module Leader: Mr. Banuka Athuraliya

IIT ID UoW ID Student Name

20210482 w1867602 Mohamed Fahim Shafna Zainab

20200730 w1833551 Munasinghe Arachchige Sandil Methmin Munasinghe

20210443 w1870513 Muhammed Azahim Muhammed Adheeb

20210729 w19128867 Mallawa Witharanage Minoli Kamwinie Dayarathne

20210129 w1870597 Melewwe Thanthrige Mewan Manodhya Amarasinghe

Student Full Name : Mohamed Fahim Shafna Zainab

Registration No : 20210482 | w1867602

Student Full Name : Munasinghe Arachchige Sandil Methmin Munasinghe

Registration No : 20200730 | w1833551

Student Full Name : Muhammed Azahim Muhammed Adheeb

Registration No : 20210443 | w1870513

Student Full Name : Mallawa Witharanage Minoli Kamwinie Dayarathne

Registration No : 20210729 | w19128867

Student Full Name : Melewwe Thanthrige Henry Mewan Manodhya Amarasinghe

Registration No : 20210129 | w1870597

Date : 13th January 2023

Keywords: Autism, Inclusive Education, Teaching, Assistive Technology, Social, Creative,

2.2 Concept Graph 15

4.4.4 Summary of Findings 60

Appendix - Section B - Methodologies XV

Figure 1 : Rich Picture Diagram of Project SupAut (Self - Composed)

Figure 31 : Collaboration between G-Drive and Slack

Table 1 : Comparison between existing products

Table 31 :Use Case descriptions for “Rephrasing”

APA American Psychiatric Association

API Application Programming Interface

ASD Autism Spectrum Disorder

BCS British Computer Society

BERT Bidirectional Encoder Representations from Transformers

CDC Centers for Disease Control and Prevention

DSM Diagnostic and Statistical Manual of Mental Disorders

FAST Families and Schools Together

GLIDE Guided Language to Image Diffusion for Generation and Editing

GPT Generative Pre-trained Transformer

HTTP Hyper Text Transfer Protocol

IEP Individualised Educational Plans

KPMG Klynveld Peat Marwick Goerdeler

NASS National Autism Spectrum Disorder Surveillance System,

NGO Non-governmental Organisations

NLP Natural Language Processing

NLTK Natural Language Toolkit

OOADM Object Oriented Analysis Design Method

PHA Public Health Agency

RNN Recurrent Neural Network

RoBERT Robustly Optimized BERT

SEN Special Educational Needs

SLEP Social, Legal, Ethical and Professional Issues

SMS Short Message Service

SRS System Requirements Specification

SSADM Structured System Analysis Design Method

UNCRPD United Nations Convention on the Rights of Persons with Disabilities

UNESCO United Nations Educational, Scientific and Cultural Organization

WBS Work Breakdown Structure

1.1 Chapter Overview

1.2 Problem Background

1.2.1 Autism And Inclusive Education