Recovering fine grained traceability links between software mandatory constraints and source code

Abstract

Software traceability is a necessary process to carry out source code maintenance, testing and feature location tasks. Despite its importance, it is not a process that is strictly conducted since the creation of every software project. Over the last few years information retrieval techniques have been proposed to recover traceability links between software artifacts in a coarse-grained and middle-grained level. In contexts where it is fundamental to ensure the correct implementation of regulations and constraints at source code level, as in the case of HIPAA, proposed techniques are not enough to find traceability links in a fine-granular way. In this research, we propose a fine-grained traceability algorithm to find traces between high level requirements written in human natural language with source code lines and structures where they are implemented.