How to Provide Developers only with Relevant Information?

After the release of a new software version it is difficult for individual developers to keep track of all newly submitted bug reports complicating their decision making, e.g., which bug to resolve next? This problem is emphasized by the presence of further information sources, such as social media, which offer valuable user feedback to developers regarding the software. However, due to an abundant amount of information, developers might never notice this feedback. Hence, we envision a real-time system that provides developers with relevant information for improving the quality of their system while filtering out irrelevant facts from multiple information sources. For this system to work, it is necessary to compute the similarity between different types of documents, e.g., tweets and bug reports, in order to detect whether they are relevant to a developer or not. In this feasibility study, we focus on analyzing this core assumption in a simplified scenario in which we identify related bugs for a given software fix with the help of Natural Language Processing methods. In this experimental setting, which exhibits the key characteristics of our envisioned system, we obtain promising results indicating that our approach is feasible.