Scribd Logo
Upload

Welcome to Scribd!

  • 25 views

    A Software Defect Prediction System Using Cohesion Metrics: Click To Edit Master Subtitle Style Under The Guidance of

    Uploaded by

    Chandrakanth Reddy
    This document proposes a software defect prediction system using conceptual cohesion metrics. It discusses existing approaches that rely on structural source code information and proposes a new metric called C3 that measures cohesion using unstructured information like comments and identifiers. C3 is calculated using Latent Semantic Indexing on the text corpus created from the source code. The architecture involves modules for computing existing LOCM and LCOM metrics as well as the new C3 metric, and comparing cohesion results. The goal is to improve defect prediction by combining structural and conceptual metrics.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as PPTX, PDF, TXT or read online from Scribd

    A Software Defect Prediction System Using Cohesion Metrics: Click To Edit Master Subtitle Style Under The Guidance of

    Uploaded by

    Chandrakanth Reddy
    25 views29 pages
    This document proposes a software defect prediction system using conceptual cohesion metrics. It discusses existing approaches that rely on structural source code information and proposes a new metric called C3 that measures cohesion using unstructured information like comments and identifiers. C3 is calculated using Latent Semantic Indexing on the text corpus created from the source code. The architecture involves modules for computing existing LOCM and LCOM metrics as well as the new C3 metric, and comparing cohesion results. The goal is to improve defect prediction by combining structural and conceptual metrics.

    Original Title

    sdpucc

    Copyright

    © Attribution Non-Commercial (BY-NC)

    Available Formats

    PPTX, PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    This document proposes a software defect prediction system using conceptual cohesion metrics. It discusses existing approaches that rely on structural source code information and proposes a new metric called C3 that measures cohesion using unstructured information like comments and identifiers. C3 is calculated using Latent Semantic Indexing on the text corpus created from the source code. The architecture involves modules for computing existing LOCM and LCOM metrics as well as the new C3 metric, and comparing cohesion results. The goal is to improve defect prediction by combining structural and conceptual metrics.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as PPTX, PDF, TXT or read online from Scribd
    Download as pptx, pdf, or txt
    25 views29 pages

    A Software Defect Prediction System Using Cohesion Metrics: Click To Edit Master Subtitle Style Under The Guidance of

    Uploaded by

    Chandrakanth Reddy
    This document proposes a software defect prediction system using conceptual cohesion metrics. It discusses existing approaches that rely on structural source code information and proposes a new metric called C3 that measures cohesion using unstructured information like comments and identifiers. C3 is calculated using Latent Semantic Indexing on the text corpus created from the source code. The architecture involves modules for computing existing LOCM and LCOM metrics as well as the new C3 metric, and comparing cohesion results. The goal is to improve defect prediction by combining structural and conceptual metrics.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as PPTX, PDF, TXT or read online from Scribd
    Download as pptx, pdf, or txt
    of 29

    A software defect prediction system using cohesion metrics

    Click to edit Master guidance of Under the subtitle style


    Mrs. K.L.S. Soujanya, Associate professor

    By
    B Chandrakanth Reddy K Sharath 4/28/12 P Krishna Kanth

    contents
    Introduction Objective Problem statement Existing system Proposed system Architecture Requirements Output screens
    4/28/12

    introduction
    Software defect prediction
    Software defect prediction is an economically

    important activity and so has received a good deal of attention.

    Cohesion
    High cohesion is a desirable property of

    software as it positively impacts understanding, reuse, and maintenance.

    4/28/12

    introduction
    Software cohesion
    cohesionis

    a measure of how stronglyrelated each piece of functionality expressed by the source code of a softwaremodule is.

    Cohesion
    Software cohesion can be defined as a

    measure of the degree to which elements of a module belong together


    Cohesion is usually measured on structural

    information extracted solely from the source code.


    4/28/12

    objective
    Existing approaches are largely based on

    using the structural information from the source code, such as attribute references, in methods to measure cohesion. classes in OO software systems based on the analysis of the unstructured information embedded in the source code, such as comments and identifiers.

    Propose a new measure for the cohesion of

    the new measure with an extensive set of

    existing metrics and uses them to construct models that predict software faults. 4/28/12

    Problem statement
    Existing system
    Existing approaches are largely based on

    using the structural information from the source code, such as attribute references, in methods to measure cohesion. information extracted solely from the source code Logical Relatedness of Methods (LORM) and the Lack of Conceptual Cohesion in Methods (LCOM)
    4/28/12

    Cohesion is usually measured on structural

    existing cohesion metrics for OO software, the

    Existing system
    Disadvantages
    Lacking of high cohesion. Lacking of measurement in cohesion. It fails to check weather the class is concept

    oriented or not.

    4/28/12

    Proposed system
    Proposed system
    We are proposing a new metric called C3, that

    measures the cohesiveness of a source file using the un-structural information from it. is retrieved from the source code like comments and identifiers.

    In proposed System unstructural information

    Information is retrieved from the source code

    using Latent Semantic Indexing.


    With the help of C3 and existing metrics ,we

    are comparing both the results and use them to construct models that predict4/28/12 defects.

    Proposed system
    Advantages
    Effective measurement of cohesion for a class. Check weather the class is concept oriented or

    not.
    Calculating the cohesion for object oriented

    systems in effective way.

    4/28/12

    Proposed metric Conceptual Cohesion of Classes the conceptual aspects of (C3) Captures
    class cohesion, as it measures how strongly the methods of a class relate to each other conceptually.
    The conceptual relation between

    methods is based on the principle of textual coherence


    C3 is based on the analysis of textual

    information in the source code, expressed in comments and 4/28/12 identifiers.

    Latent Semantic Indexing (LSI)


    We use LSI to extract and analyze the

    unstructured information from the source code

    LSI is an advanced IR method Reasons for using LSI


    concept and feature location traceability link recovery between source code

    and documentation
    identification of abstract data types in legacy

    source code
    clone detection
    4/28/12

    LSI
    Measuring Text Coherence with Latent

    Semantic Indexing

    Compute the similarities between consecutive

    sentences in the text


    From Textual Coherence to Software

    Cohesion

    extraction of relevant information from the

    source code i.e. all identifiers and comments from the source code 4/28/12

    architecture

    4/28/12

    modules
    LOCM measure computation
    We measure the cohesive nature of source file

    using structural information

    C3 measure computation
    Using C3 metrics we compute the cohesion

    strength of source class using LSI IR technique on un- structural information

    LCOM and C3 cohesive comparison

    designers module

    4/28/12

    LCOM computation
    Here we compute the cohesion measure of

    source file using its structural information like variables, methods etc
    Input a preprocessed source file

    Steps to compute

    Extract structural information from the source Apply the LCOM formula on each method of

    source
    If the result is 0,then we say that the class is

    highly cohesive
    4/28/12 If the result is 1,then it is less cohesive

    Conceptual Cohesion of Classes(C3) computation


    A tool IR-based Conceptual Cohesion Class

    Measurement, which supports this methodology and automatically computes C3 for any class in a given software system preprocessed and parsed to produce a text corpus. Comments and identifiers from each method are extracted and processed. A document in the corpus is created for each method in every class.

    Corpus creation: The source code is

    Corpus indexing: LSI is used to index the

    corpus and create an equivalent semantic space. 4/28/12

    C3 measure

    4/28/12

    C3 computation
    consider an OO system as a set of classes

    C ={ c1; c2 . . . cn }.
    The total number of classes in the system C is

    n = mod(C).
    A class has a set of methods. For each class c E

    C, M(c)={m1; . . .;mk} is the set of methods of class c.

    Conceptual Similarity between Methods(CSM)

    4/28/12 where vmk and vmj are the vectors corresponding to the

    mk;mj E M(ci) methods, T denotes the transpose, and

    C3 computation
    Average Conceptual Similarity of Methods

    in a class (ACSM), The ACSM c E C is

    (C3). For a class c 2 C, the conceptual

    cohesion of c, C3(c) is defined as follows:

    4/28/12

    conclusion

    The combination of structural and

    conceptual cohesion metrics defines better models for the prediction of faults in classes than combinations of structural metrics alone.

    4/28/12

    System Requirements
    Hardware Requirements:

    Processor:Intel Duel core Hard Disk :60GB Monitor :LCD color RAM :512 MB
    Software Requirements:

    Operating System

    4/28/12

    output

    4/28/12

    Compile all the java files

    4/28/12

    Browse a source file

    4/28/12

    LCOM cohesive measure

    4/28/12

    Browse file for c3 measure

    4/28/12

    Result of C3 measure

    4/28/12

    Output in cmd

    4/28/12

    Thank you
    4/28/12

    You might also like