A model for conformance analysis of software documents

During the evolution of a large-scale software project, developers produce a large variety of software artifacts such as requirement specifications, design documents, source code, documentation, bug reports, etc. These software documents are not isolated items - they are semantically related to each other. They evolve over time and the set of active semantic relationships among them is also dynamic. Their evolution makes the task of managing them and maintaining their semantic consistency a very challenging task for software developers. At times, the evolutionary changes may reduce the consistency of the software project and break semantic connections among its documents. We use the term conformance to denote the state where the network formed by software documents and their relationships is in semantic harmony. Conformance analysis is the process of determining whether software documents and their logical relationships are in agreement. We present a representation for software documents and their logical relationships based on the hypertext model. We describe how conformance analysis can be supported by this representation and present a method to detect nonconformance using timestamps and versioned hypermedia. Then we describe a formalism, called the conformance model, that can be used to understand and combine approaches to the conformance analysis problem.