A

MINI PROJECT REPORT

ON
INTELLIGENT SPELLING CORRECTOR
BE(IT)-IV SEM
By
V. Naga Madhuri (160121737023)
D. Sushanthi (160121737304)
Under the guidance of
Mr.Saketh Kallepu
IT Department

DEPARTMENT OF INFORMATION TECHNOLOGY

CHAITANYA BHARATHI INSTITUTE OF TECHNOLOGY (A)
(Affiliated to Osmania University; Accredited by NBA(AICTE) and NAAC(UGC),
ISO Certified 9001:2015)

KOKAPET(V), GANDIPET(M), RR District HYDERABAD – 75

Website: https://www.cbit.ac.in/

2022-2023
 CERTIFICATE

This is to certify that the project seminar report entitled “INTELLIGENT SPELLING
CORRECTOR” submitted to CHAITANYA BHARATHI INSTITUTE OF
TECHNOLOGY, in partial fulfilment of the requirements for the award of the
completion of V semester of B.E in Information Technology, during the academic year
2023-2024, is a record of original work carried out by VANAM NAGA MADHURI
(160121737023) DEVARAJU SUSHANTHI (160121737304) during the period of study
inDepartment of Information Technology, CBIT, Hyderabad.

Project Guide Head of Department

Mr.Saketh Kallepu Dr. A Rajanikanth

Professor, Dept. Of IT,

CBIT, Hyderabad.
 ACKNOWLEDGEMENTS

In this project, we picked up and used a lot of new technical abilities at the professional level. But
without the assistance of other people, it would not have been possible. I want to give each one of
them my deepest appreciation.

We would like to convey our deep gratitude and appreciation to our mentor, Saketh Kallepu sir,
for her leadership, persistent monitoring, and provision of the relevant details concerning this
project. We appreciate Dr. A. Rajanikanth, our department head, for providing the resources
needed to complete this project as well as his consistent support.

We also want to express our gratitude to our parents and friends for their wonderful support and
inspiration, which assisted us in the project's development. Finally, we would like to express our
gratitude to CBIT, our university, for giving us the chance to work on this project.
ABSTRACT
A sophisticated software program called an intelligent spelling corrector is made to automatically
identify and correct spelling mistakes in written text. These systems tokenize incoming text, look
for possible misspellings, and produce a list of suggested repairs by utilizing complex algorithms
and natural language processing (NLP) concepts. Then, these candidates are graded according to
how likely it is that they would be the appropriate replacement, frequently taking context into
account to increase accuracy. Edit distance algorithms, n-gram language models, neural networks,
and statistical language models may all be used in the correction process. The fundamental
processes and strategies for developing intelligent spelling correctors are summarized in this
abstract, emphasizing the significance of precise language models and sufficient training data for
these systems' efficacy.

Keywords:

• Contextual analysis

• Tokenization

• Natural language processing (NLP)

• Spelling Correction

• Error detection

5
TABLES OF CONTENTS

S.No Topics Page No.

1 CHAPTER 1
8
1.1 Introduction

1.2 Problem Statement 9

1.3 Objective of the Project 10

2 CHAPTER 2

2.1 Literature Survey

11
12
2.2 Existing System

3 CHAPTER 3

3.1 Proposed Methodology and System

13

3.2 System Specifications 14

4 CHAPTER-4

4.1 Implementation and Description 15-19

5 CHAPTER-5

1. Result and Output 20-22

6 CHAPTER -6

1. Conclusion 23
2. Future Scope 24

7 References 25

6
 TABLE OF FIGURES

S. No Figure Name Figure No. Page No.

1 Fig 5.1.1. 20
Home Page

2 After clicking on Correct Fig:5.1.2 21

Spellings button

3 21
List of possibly misspelled Fig:5.1.3
words

4 22
Candidate Words Fig:5.1.4

7
 CHAPTER -1
1.1 INTRODUCTION

In the age of digital communication and information exchange, accurate spelling plays a crucial
role in conveying our thoughts and ideas effectively. However, as humans, we are prone to making
spelling errors, be it due to haste, oversight, or simply the complexity of the English language.
These errors can mar the clarity and professionalism of our written content, whether it's in emails,
reports, essays, or social media posts. Intelligent spelling correctors have revolutionized the way
we engage with written language to address this prevalent difficulty. An advanced piece of
software called an "intelligent spelling corrector" goes beyond the standard red squiggly lines we
see in word processors to automatically identify and correct spelling errors in text. In contrast to
their predecessors, these modern spell checkers use artificial intelligence, machine learning, and
extensive language databases to offer users precise and context-sensitive spelling suggestions. An
essential tool for authors, professionals, students, and anybody who communicates through written
language in this era of AI-driven language technology is the intelligent spelling corrector. This
introduction explores the development of spelling correction technology, its amazing capabilities,
and how it might improve the value and effect of written communication across a range of fields.
We will examine the inner workings of these intelligent systems, their applications, and the
advantages they offer consumers across the board in the discussion that follows. These tools can
adapt to user preferences and writing styles over time. They may learn from user interactions,
incorporating user-specific spelling choices into their suggestions for greater accuracy. Intelligent
spell checkers frequently support many platforms, enabling customers to use their spelling
correction features on a variety of gadgets, including PCs, smartphones, and tablets.

8
1.2 PROBLEM STATEMENT
Written communication is essential in today's digital world for personal, professional, and
academic contexts. However, a significant barrier to efficient written communication continues to
be the prevalence of spelling mistakes. Although there are traditional spell-checking systems
available, their efficiency is limited by their inability to comprehend context and change to shifting
linguistic patterns. Because of this, people in a variety of fields still struggle with spelling mistakes
that lessen the impact, professionalism, and clarity of their written content.The primary problems
to be solved in this project are as follows:

1. Prevalence of Spelling Errors: Spelling errors are a common issue in written communication,
affecting personal, professional, and academic contexts.

2. Limitations of Conventional Tools: Traditional spell checkers lack the ability to grasp context
and adapt to changing language trends, rendering them ineffective in addressing the root causes of
spelling errors.

3. Complexity of English Language: The intricate nature of the English language, characterized
by a vast vocabulary and nuanced contextual meanings, poses a significant challenge for accurate
spelling correction.

4. Need for Context-Aware Solutions: There is a growing demand for intelligent spelling
correctors capable of not only identifying misspelled words but also comprehending the context in
which they appear.

5. User-Centric Adaptation: Effective solutions should adapt to the individual writing style of
users, offering personalized suggestions and real-time feedback as they type.

The spelling errors are a common hindrance, detracting from clarity and professionalism.
Conventional spell-checking tools, lacking contextual awareness, struggle to address this issue
effectively. To overcome this challenge and improve written communication, there's a pressing
need for intelligent spelling correction solutions capable of context-sensitive suggestions and
adapting to evolving language trends.

9
1.3 OBJECTIVE OF THE PROJECT

The objective of an intelligent spelling corrector is to automatically detect and rectify spelling
errors in text to improve the accuracy and readability of written content. This technology is
designed to:

1. Enhance Communication: Spelling errors can hinder effective communication. An intelligent

spelling corrector helps users convey their intended message more clearly by fixing typos and
misspellings.

2. Improve Professionalism: In business and academic contexts, polished and error-free writing is
essential for conveying professionalism and competence. A spelling corrector can help ensure that
written documents maintain a high standard.

3. Enhance User Experience: Many software applications, websites, and text editors integrate
spelling correction to provide a better user experience. Users expect a certain level of error
correction when using these tools.

4. Support Accessibility: Spelling correctors can be invaluable for individuals with dyslexia,
learning disabilities, or language barriers, as they help ensure that written content is more
accessible.

5.Maintain Consistency: In technical or scientific writing, consistent spelling is crucial to avoid

confusion or misinterpretation. A spelling corrector can help maintain this consistency.

6.Handle Context: Advanced spelling correctors consider the context in which words appear to
provide more relevant suggestions. This includes understanding grammar, syntax, and the meaning
of nearby words.

7.Suggest Alternative Words: Instead of just fixing misspelled words, intelligent spelling
correctors can suggest alternative words or phrases to improve the overall quality of the text.

Overall, the primary objective of an intelligent spelling corrector is to enhance the quality,
accuracy, and clarity of written communication while simplifying the process of error detection
and correction for users.

10
 CHAPTER-2

2.1 LITERATURE SURVEY

“Adaptive Context-Sensitive Spelling Error Correction Techniques for the Extremely

Unpredictable Error Generating Language Environments”

Authors: Minho Kim, Sung-Ki Choi, Jingzhi Jin, Hyuk-Chul Kwon

Spelling mistakes are separated into two categories: context-sensitive spelling mistakes and real
word spelling mistakes. Nonword errors are those that involve language that is not found in
dictionaries. Only by carefully studying the text's morphology can such a mistake be found.
However, when left and right context are considered, context sensitive spelling errors appear to be
correct when seen as a word unit. Since the relevant vocabulary can only be determined by taking
the context's meaning and the syntactic relationship into account, this has a very high correction
difficulty. There are two categories of context-sensitive spelling error correction: rule-based
correction methods and statistical information-based correction methods. It is practically hard to
create rules that accurately reflect all linguistic events that occur in the actual world, hence rule-
based correction methods require the expert to be equipped with high level linguistics and
computer science knowledge. Formulaic or errors that occur frequently can be corrected with a
high degree of probability using a rule-based approach, but errors that arise due to entry errors
cannot be corrected using merely a rule-based approach and are far more difficult to repair. This
study recommended a corrective technique based on statistical data. The model that made use of
the research's corrected vocabulary pair bears the restriction that it can only be used with the
specific vocabulary pair that was chosen for the study. This research proposed a noisy channel
model-based statistical context-sensitive spelling error detection and repair model that focused on
the entire eojeol to get around this constraint.

11
2.2 EXISTING SYSTEM

Numerous systems and software programmers now in use have intelligent spelling correcting
functions. Remember that since then, other systems may have appeared because the field of natural
language processing (NLP) is continually developing. Here are some examples of existing
intelligent spelling corrector systems:

Microsoft Word Spell Checker: Microsoft Word includes a built-in spelling and grammar
checker that suggests corrections for misspelled words and grammatical errors as you type. It uses
a combination of rule-based and statistical approaches.

Google Docs Spell Check: Google Docs provide real-time spelling and grammar checking while
you write. It also offers suggestions for corrections and provides context-aware recommendations.

Grammarly: Grammarly is a popular writing assistant tool that offers spelling correction,
grammar checking, and style suggestions. It works as a browser extension, desktop app, and can
be integrated into various writing applications.

Hunspell: Hunspell is an open-source spell checker and morphological analyzer library used in
various software applications. It provides dictionaries for multiple languages and is used in
applications like LibreOffice and Mozilla Firefox.

Language Tool: Language Tool is an open-source proofreading software that offers spell and
grammar checking. It supports multiple languages and can be used as a standalone application or
integrated into other tools.

ProWritingAid: ProWritingAid is a writing analysis tool that provides spelling and grammar
checking, style suggestions, and in-depth reports on writing improvement. It can be used online or
as a desktop application.

AutoCorrect: Many mobile keyboards, including the default iOS and Android keyboards, feature
autocorrect functionality. These keyboards automatically suggest and correct spelling errors as you
type.

12
 CHAPTER-3
3.1 PROPOSED METHODOLOGY AND SYSTEM
It involves a structured approach, beginning with the collection of a diverse and substantial corpus
of text data encompassing both correct and misspelled words and phrases. Following this, data
preprocessing is performed to clean and tokenize the text, making it ready for further analysis.
Feature engineering is employed to extract relevant linguistic features, while machine learning
model selection involves choosing an appropriate algorithm, such as statistical models, deep
learning networks, or rule-based systems, for the spelling correction task. Fine-tuning and iterative
refinement are then conducted to enhance accuracy. In terms of system design, user-friendly
interfaces are developed to facilitate user interactions. The spelling correction engine, powered by
the trained model, detects, and corrects spelling errors, offering suggestions, and considering
contextual information. User feedback is integrated to improve future predictions. Customization
options, scalability and security considerations are also incorporated, ensuring a robust and user-
centric intelligent spelling corrector system.

Proposed Features:
• Gather a diverse and substantial dataset of text, including both correctly spelled words and
misspelled words or phrases. This dataset will serve as the foundation for training and
testing the spelling correction model.
• Extract relevant features from the text data. Common features include word frequencies,
character n-grams, contextual information (e.g., neighboring words), and semantic
embeddings.
• Clean and preprocess the text data by removing special characters, punctuation, and
unnecessary formatting. Tokenize the text into words or subword units, and convert all text
to a consistent case (e.g., lowercase).
• Divide the preprocessed dataset into training, validation, and test sets. The training set is
used to train the spelling correction model, the validation set helps tune hyperparameters,
and the test set is reserved for final evaluation.
• Integrate the trained spelling correction model into the target application or system where
it will be used. Ensure it can efficiently process text input and provide real-time corrections.

13
 3.2 SYSTEM SPECIFICATIONS

3.2.1 HARDWARE REQUIREMENTS

Computer hardware is a collective term used to describe any of the physical components of an
analogue or digital computer. Computer hardware can be categorized as being either internal or
external components. Generally, internal hardware components are those necessary for the
proper functioning of the computer, while external hardware components are attached to the
computer to add or enhance functionality. A list of hardware required is given below.

• Processor: Intel dual Core or higher

• RAM: 8 GB or higher
• Hard Disk: 256 GB or higher
• Operating System: Windows XP or higher

3.2.2 SOFTWARE REQUIREMENTS

The software requirements for developing and running Intelligent Spelling Corrector are as
follows:

• Development Environment: Visual Studio Code ,PyCharm , Jupyter Notebook or any

other preferred one.
• Front-end: HTML, CSS,
• Back-end :Flask
• Programming Languages: Python

14
 CHAPTER-4

1. IMPLEMENTATION AND DESCRIPTION

This Python script processes an uploaded file, performs text correction, and identifies potentially
misspelled words. After uploading and reading the file, it utilizes the TextBlob library to correct
the text, attempting to fix any spelling errors. To prepare the text for spell checking, it removes
punctuation marks. Then, it employs the PySpellChecker library to find words in the cleaned text
that might be misspelled, storing them in a list. For each potentially misspelled word, it determines
the most likely corrected version using the SpellChecker, adding it to one list, and compiles a list
of likely correction candidates for further reference. This script is a practical tool for correcting
text and detecting spelling mistakes in uploaded files, providing both the most probable corrections
and additional alternatives for potentially erroneous words. Make sure to have the TextBlob and
PySpellChecker libraries installed for it to function correctly.

15
This HTML code snippet depicts a form designed to facilitate the upload and correction of text
files. Enclosed within a <div> element styled with the "form-group" class, the form consists of
several components. The <h2> element serves as a potential title or header, though it currently
lacks visible text. The <form> element is pivotal to the form's functionality, set to submit data
using the POST method and configured to handle file uploads through the "multipart/form-data"
enctype attribute, crucial for transmitting binary data, such as files, to the server.

Beneath the form element, a <p> element, styled with the "font-weight-bold" class, provides clear
instructions to users, directing them to upload a text file with a ".txt" extension that requires
spelling correction. An <input> element with the "file" type allows users to select a file from their
local device, with the "name" attribute set as "file" for identification during form submission.
Finally, another <input> element, styled as a button with various classes, including "btn," "btn-
primary," "font-weight-bold," and "text-white," bears the label "Correct Spellings." When users
click this button, it triggers the form submission process, sending the chosen file to the server for
spell checking. In summary, this form provides a user-friendly means to upload text files for
convenient spelling correction with a single click.

This HTML code segment represents a section for displaying the original text and the corrected text,
typically used in a web application or user interface after a spell-checking or text correction process has
taken place. Here's a breakdown of its components:Enclosed within an element, the section begins with an
element labeled "Text to be corrected." This heading serves as a clear title, indicating that the following
content will display the original text that requires correction.

16
Next, there is an element with several classes applied: "font-italic," "lead," and "text-justify." This
paragraph is populated dynamically with text content from a variable called a. The use of double curly
braces suggests that this text is likely pulled from a data source or generated by a server-side script. It's
styled with italics and justified text alignment, enhancing readability.

Following the original text display, another element appears with the label "Corrected Text." This heading
signals that the content following it will showcase the text after spelling correction has been applied.

The final element, similar to the previous one, uses classes such as "font-italic," "lead," and "text-justify."
It displays the corrected text, which is dynamically inserted from a variable called correct. Like the original
text, the corrected text is also styled with italics and justified alignment for consistency.

In summary, this HTML section is designed to visually present both the initial text and the corrected text
to users, making it easy for them to compare the changes made during the correction process. It uses HTML
headings and styling classes to organize and format the content for improved readability and clarity.

This HTML code snippet serves the purpose of displaying a list of potentially misspelled words on a web
page, commonly used in applications such as spelling correction tools. It is neatly structured within a <div>
container. A descriptive subheading, <h3>, labeled "List of possibly misspelled words," introduces the list,
making it clear to users what they can expect. The list itself is created as an ordered list (<ol>), signifying
that the items will be presented with numbers for easy reference.

To populate the list, a loop is implemented using templating syntax ({% for i in range(0, len) %}). This loop
iterates through a range of numbers, likely corresponding to the number of misspelled words to be
displayed. Within each iteration, an <li> (list item) element is generated. These list items are assigned two
classes, "list-group-item" and "list-group-item-action," which are often used for styling purposes, ensuring
a visually appealing and user-friendly design.

The content inside each list item, denoted as {{misspelled[i]}}, is dynamically generated and retrieved from
a variable named misspelled. This variable is presumed to contain an array or list of words that are

17
potentially misspelled. By looping through this list and displaying the words, the HTML code effectively
presents users with a clear and structured list of words that may require spelling correction. This user-
friendly format aids in the identification and review of potentially problematic words, making it a valuable
feature in spelling correction and text analysis applications.

This HTML code snippet is designed to present a table on a web page, displaying pairs of words:
correct words and candidate words associated with potentially misspelled words. The surrounding
<div> element ensures that the table is responsive and adjusts well on medium-sized screens. An
informative <h3> heading introduces the purpose of the table, indicating that it will show both
correct and candidate words for potential misspellings.

Within the table, the structure is divided into a table header (<thead>) and a table body (<tbody>).
The table header has a dark background color for visual contrast and clarity. It contains two
columns, labeled "Correct Word" and "Candidate Words," serving as headers for the data that
follows.

18
The data itself is generated dynamically using templating. A loop, indicated by {% for j in range(0,
len1) %}, iterates through a range of numbers. Inside each iteration, a new table row (<tr>) is
created, containing two table cells (<td>) that display the correct word and a list of candidate
words. These cells are populated with data from variables list1 and list2, respectively, suggesting
that the correct words and their corresponding candidate words are dynamically retrieved and
displayed for each iteration of the loop.

Overall, this HTML structure offers a clear and organized way to present pairs of correct and
candidate words for users to review, typically in the context of spelling correction or text analysis.
The table's styling and responsive design enhance the user experience, making it easy to examine
potential corrections for misspelled words.

So, the CSS rule you've provided instructs that any HTML element with the class "maincon"
should have a background color of "#15435c." This will give the elements a background color that
corresponds to the specified shade of blue-green.

19
 CHAPTER-5
5.1 RESULT
HOME PAGE:

Fig:5.1.1 Home page

The home page serves as the central hub of the application, providing the user with a dashboard
that showcase to choose a a file with mistakes which is in a format of .txt .Choose a file and click
on correct spellings button.It serves as the starting point users to with the content of the webpage.

20
 Fig:5.1.2 After clicking on Correct Spellings button

The above page will be found if we click on the Correct Spellings button.We get the Text to be
corrected and correct text which we have sent in the .txt file.In real-world spelling correction
systems can be much more complex. You can enhance this system by considering factors such as
context and word frequency.

Fig:5.1.3 List of possibly misspelled words

21
The list of possible misspelled words say the errors which we got in the file which we have
checked.Creating a comprehensive list of possible misspelled words for a spelling corrector is a
challenging task due to the vast number of potential misspellings in any language.

Fig:5.1.4 Candidate words

In a spelling corrector our website, candidate words are potential replacement words suggested to
the user when a misspelled or incorrectly typed word is detected. These candidate words are
generated based to provide the user with a list of possible corrections.

22
 CHAPTER-6

6.1 CONCLUSION
The development and implementation of an intelligent spelling corrector represent a significant
advancement in language processing technology. Such systems have the potential to greatly
enhance written communication across various domains. By leveraging sophisticated algorithms,
artificial intelligence, and large language models, these spelling correctors can detect and rectify
errors with remarkable accuracy. The benefits of intelligent spelling correctors are manifold. They
aid in improving the overall quality of written content, making it more professional and easier to
understand. Additionally, they assist individuals with varying levels of literacy and language
proficiency, fostering inclusivity and accessibility. These systems can also be invaluable tools for
students, professionals, and writers who rely on error-free text.

However, challenges remain, such as the potential for overcorrection or misinterpretation of

creative or specialised language. Privacy and data security concerns also need to be addressed,
given the sensitivity of the textual data they process. While they have made tremendous progress,
there is always room for improvement, particularly in understanding and adapting to regional and
domain-specific language variations. Intelligent spelling correctors are a testament to the
capabilities of AI and NLP technologies, offering substantial advantages in terms of language
accuracy, accessibility, and efficiency. Their continued refinement and integration into various
applications hold great promise for the future of written communication.

23
6.2 FUTURE SCOPE

• Enhanced Contextual Understanding: Future spelling correctors will increasingly incorporate

advanced contextual analysis to provide more accurate suggestions based on the context of the
sentence. This will reduce the chances of misinterpretation and improve overall text quality.

• Customisation and Personalisation: Spelling correctors will allow users to customise and
personalise their correction preferences. Users may choose different levels of formality, industry-
specific terminology, and even adapt to regional language variations.

• Machine Learning Advancements: Ongoing advancements in machine learning techniques will

continue to improve the accuracy and efficiency of spelling correctors, reducing false positives and
negatives.

• Mobile and IoT Integration: Integration into mobile devices, IoT devices, and smart assistants
will make spelling correction more seamless across various platforms and applications.

• Content Generation and Automation: Advanced spelling correctors will play a crucial role
in content generation, aiding writers in generating error-free content more efficiently.

24
 CHAPTER-7

7.1 REFERENCES

• IEEE (Used to explore other projects): https://ieeexplore.ieee.org/document/9630371

• Ginger Software: https://www.gingersoftware.com/spellcheck
• Grammar: https://www.grammar.com/intelligence-correct-spelling
• Language Tool: https://languagetool.org/
• IRJET: https://www.irjet.net/archives/V6/i6/IRJET-V6I674.pdf
• Trinka: https://www.trinka.ai/spell-checker
• Cambridge university: https://www.cambridge.org/core/journals/natural-language-
engineering/article/abs/arabic-spelling-error-detection-and-
correction/630859B3EC3365926DBAFF92436A89A6
• IELTS: https://ieltsetc.com/ielts-essay-checker/
• Nuijten, M. B., Hartgerink, C. H. J., van Assen, M. A. L. M., Epskamp, S. & Wicherts, J.
M. Behav. Res. Methods 48, 1205–1226 (2016).
• Nuijten, M. B., van Assen, M. A. L. M., Hartgerink, C. H. J., Epskamp, S. & Wicherts, J.
M. Preprint at PsyArXiv(2017)

25