This document summarizes a student project on analyzing software design defects using machine learning. The project will involve detecting various types of code smells in source code, preparing datasets of code classes, using machine learning algorithms to test for smells, and analyzing the causes and limiting criteria of detected smells. The expected outcomes are to identify the types of smells found, their causes, and limiting criteria. The document provides background on code smells and some examples. It outlines the research pathway and references literature on evaluating code smells and detection tools.

1. Mentor:-
Ms. Aakanshi Gupta
AMITY SCHOOL OF ENGINEERING AND TECHNOLOGY,
NEW DELHI- 110061
An institution of Ritnand Balved Education Foundation (RBEF)
Branch: Computer Science and Engineering (CSE)
Pre-Synopsis Presentation
Team Members:-
Kritika Phulli [01610402716]
Nidhi Kumari Chauhan [02110402716]
Deepanshu Sharma [35210402716]
Vishal Bhat [41910402716]

2. Software Design Defects
And
Its Analysis
Using Machine Learning

3. Problem
Statement
Background
Of The Study
Expected
Outcome
References
Research
Design

4. Software
Engineering
SDLC

5. Design Issues
• Caused by Sub-optimal
design decisions which can
be identified in the source
code as code smells,
negatively impact the
software quality.
• Poor design decisions.

6. Code
Smells

7. Bad Smells
The noticeable factors of code smell was
first designed by Fowler.
“Smell are symptoms of poor design or
implementation choices that causes long term
maintainability problem and mask bugs.”
Some of the bad smells are listed below :
Member Ignoring Method Slow Loop
Leaking Inner Class Unclosed
Closable.

8. Research
Pathway
DETECTION • Detection of different type of smell is
done initially.
ANALYSIS
• Preparation of dataset for
classes of the code.
ML
TESTING
• Using Machine Learning
Algorithms, Further
testing Would Be Done.

9. Expected
Outcome
• Type of smells
• Causes
• Limiting Criteria
Once Data is collected and Data sets are created,
We expect the outcomes emphasizing on :

10. References
• Thanis Paiva, Amanda Damasceno Eduardo Figueiredo and
Cláudio Sant’Anna, “On the evaluation of code smells and
detection tools”, Journal of Software Engineering Research
and Development 2017.
• A. Yamashita and L. Moonen, “Exploring the impact of inter-
smell relations on software maintainability: An empirical
study,” in Proceedings of the International Conference on
Software Engineering (ICSE). IEEE, 2018.

11. Thank You