Integration of design smells and role-stereotypes classification dataset

Area of Science:

• Software Engineering
• Computer Science

Background:

• Design smells are recurring patterns that negatively impact software maintainability.
• Role stereotypes define the responsibilities of classes in software design.
• Understanding the interplay between design smells and role stereotypes is crucial for improving software quality.

Purpose of the Study:

• To establish a fine-grained dataset integrating design smell detection and role stereotype classification.
• To investigate the relationship between design smells and role stereotypes in software systems.
• To provide a resource for analyzing how design smells manifest across different software roles.

Main Methods:

• A dataset was created from 12 open-source Java projects from GitHub.
• The dataset includes 18 design smell indicators and classifications for 2,513 Java classes across six role stereotypes.
• Unsupervised learning was used to cluster the data into ten groups, highlighting co-occurring design smells.

Main Results:

• A comprehensive dataset linking design smells and role stereotypes was systematically generated.
• The analysis identified specific clusters of design smells associated with particular role stereotypes.
• The findings illuminate previously unrecognized relationships between software design flaws and class responsibilities.

Conclusions:

• The developed dataset provides a valuable resource for researchers and practitioners in software engineering.
• The insights gained can guide efforts to mitigate design smells and improve software maintainability.
• This work contributes to a deeper understanding of software design patterns and their impact on system evolution.