Automated Traceability Techniques for Software Engineering and e-Science

The effective management of information is a fundamental task in artifact-rich domains such as e-Science and software engineering. Data provenance techniques in e-Science seek to capture the relationships between data sets, methods, and experimental results, in order to foster scientific reproducibility. Similarly, software traceability is concerned with identifying connections between related artifacts in the software engineering lifecycle. This talk highlights the parallels that exist between these two tasks, allowing for the cross-fertilization of techniques. In the past, we used insights from scientific workflows and provenance systems in e-Science to address software traceability challenges, leading to the development of a system known as Architecture-Centric Traceability for Stakeholders (ACTS). We have also combined ACTS with techniques
like topic modeling to support the automated semantic analysis of captured links. Currently, we are investigating how software traceability techniques can inform the development of data provenance systems for e-Science. We seek to lower the barrier to entry to automated data provenance and provide provenance support across heterogeneous and distributed data. The talk will conclude with an informal discussion of potential projects for building data provenance support using Microsoft technology.