Will it Smell? Predicting Bad Code Patterns using Machine Learning Techniques

Master Thesis for the master study programs offered by the Departament of Informatics, Faculty of Natural Sciences, University of Tirana.

Abstract:

This thesis aims to explore the use of machine learning techniques to predict bad code patterns, commonly referred to as code smells. Students will conduct repository mining and dataset exploration to extract relevant code patterns and train models for detecting problematic structures.

The study will assess various machine learning models and evaluate their effectiveness in identifying maintainability issues. The findings will contribute to understanding the potential of AI-driven code analysis and its impact on improving software quality.

Starting Point:

Dewangan, S., Rao, R. S., Mishra, A., & Gupta, M. (2021). A novel approach for code smell detection: an empirical study. IEEE Access, 9, 162869-162883.

Gupta, H., Kulkarni, T. G., Kumar, L., Neti, L. B. M., & Krishna, A. (2021, April). An empirical study on predictability of software code smell using deep learning models. In International conference on advanced information networking and applications (pp. 120-132). Cham: Springer International Publishing.

Sharma, T., & Kessentini, M. (2021, May). Qscored: A large dataset of code smells and quality metrics. In 2021 IEEE/ACM 18th international conference on mining software repositories (MSR) (pp. 590-594). IEEE.