Information retrieval models for recovering traceability links between code and documentation

Abstract

The research described in this paper is concerned with the application of information retrieval to software mainte-nance, and in particular to the problem of recovering trace-ability links between the source code of a system and its free text documentation. We introduce a method based on the general idea of vec-tor space information retrieval and apply it in two case studies to trace C++ source code onto manual pages and Java code onto functional requirements. The case studies discussed in this paper replicate the studies presented in references [3] and [2], respectively, where a probabilistic information retrieval model was applied. We compare the results of vector space and probabilistic models and formu-late hypotheses to explain the differences. 1.