Quality Improvements for Trace Links between Source Code and Requirements

[Context and Motivation] Traceability between source code and requirement artifacts is important for various tasks during software development. However, it is a lot of effort to create and maintain traceability links manually. Therefore, semi-automatic traceability support is developed. [Question/ problem] Traceability research has a strong focus on trace link recovery using information retrieval (IR) techniques. These techniques use the textual similarity of documents to create trace links. The quality according to precision and recall of these techniques is still not satisfying. [Principal ideas/ results] Precision and recall can be improved by providing more data as used in IR. In this thesis, we evaluate two new data source types to create trace links. To link two specific artifacts, we exploit existing links in the context of these artifacts. Furthermore, we use the interaction logs of developers for trace link creation between the artifacts the developers touched. [Contribution] In this paper, we present the research problems as well as the principal solutions to deal with these challenges, our research methodology, and our progress so far.