Using information retrieval to support software maintenance tasks

This paper presents an approach based on information retrieval (IR) techniques for extracting and representing the unstructured information in large software systems such that it can be automatically combined with analysis of program dependencies and execution traces to define new techniques for feature location, impact analysis, and software measurement tasks. We expect that these new techniques will contribute directly to the improvement of design of incremental changes and thus increased software quality and reduction of software maintenance costs. The presented results are based on the author's doctoral dissertation.