Master of Information Systems
School of Economics and Information Systems - Faculty of Commerce
Chen, Zerong, Semantic markup for inconsistency handling in requirements engineering, M.Info.Sys. thesis, School of Economics and Information Systems, University of Wollongong, 2004. http://ro.uow.edu.au/theses/223
Requirements Engineering (RE) involves carefully devising descriptions of systems that we seek to build, which include descriptions of requirements that such systems must satisfy, as well as descriptions of the environment in which such systems will be situated. Research in RE seeks to understand how such descriptions might be generated, what the appropriate languages are for representing such descriptions, how changes to such descriptions might be managed and support might be provided for ensuring that such descriptions satisfy several normative quality criteria. A key pre-requisite for automated, or semi-automated, tool support for RE process is the ability to make these descriptions machine-processible. This is especially important in the context of devising tool support for the difficult process of requirements inconsistency handling. The key thesis presented in this document is that the notion of semantic markup (as devised in the Semantic Web initiative), when applied to requirements specifications in semi-formal or informal notations, can form the basis for automated, or semi-automated, tool support for inconsistency handling. In support of this thesis, we present several examples of extracting machine-processible descriptions of semi-formal and informal requirements specifications. We present some general methodological guidelines for how this might be achieved and present the outlines of a new generation of end-user markup tools, which, embedded in CASE environments, might ease the process of annotating specifications with such markup. We then present the design and implementation of a system that detects (and in part, helps resolve) inconsistencies in such specifications. We believe that this can form the basis for a new research program in requirements engineering exploring the deployment of semantic markup techniques in specifications. An initial version of our work has been reported in [Chen et al., 2003}.