The objective of this project is twofold. First, we want to define methodologies for extracting the characteristics of the developers who worked on the artifacts that RSSEs analyze. Achieving this goal would allow defining proper developers’ profiles and extract novel features that can be used to improve the effectiveness of state-of-the-art RSSEs. Second, we want to provide developers with a comprehensive platform through which they can: (i) customize their experience with RSSEs using End-User Development, thus making them adaptable in terms of functionalities to be used, visualizations techniques, and other aspects; (ii) receive explanations for the recommendation provided by RSSEs, with the aim of increasing developers' trust in the decisions. We will test such novel methodologies and define three novel profile-based and developer-centered RSSEs: (i) a personalized code generation approach that takes into account the developers' coding style; (ii) a defect prediction model that considers the proneness of a developer in introducing specific types of defects; (iii) a code readability prediction approach that will provide predictions based on the developer who reads the code.
The project started on the 28th September 2023 and lasts two years.