Towards automating traceability: An incremental and scalable approach.

摘要

Software engineers face significant challenges when creating and maintaining traceability information. Many software tools lack the ability to automatically create semantic relationships between a software project's heterogeneous artifacts. Thus, traceability typically involves laborious, manual procedures and, as such, these important techniques face serious adoption hurdles by developers.; In this work, we describe a new approach to traceability that automates the creation and maintenance of traceability information among heterogeneous software artifacts. Furthermore, our approach allows these relationships to be viewed in the tool that originally created the artifact rather than forcing the user to switch to a separate tool to view these relationships. Our conceptual framework, TraceM, incorporates techniques from open hypermedia and information integration to provide scheduling, query, relationship mapping, evolution, and export services.; We have developed a prototype implementation of TraceM to evaluate our approach. Our evaluation includes a small study of its usability and utility as well as a separate scalability test using artifacts obtained from industry.