Evaluation Framework for Semantic-Enabled Software Engineering

Abstract

For years software engineering community is striving to formulate software development methodologies for the development of quality products. There are many methodologies proposed for this aim of theirs but still these method- ologies have limitations and gaps when it comes to complex softwares. Soft- ware Development methodologies faces communication and coordination is- sues, issues related to knowledge management etc. which are very funda- mental in their nature. So new technologies have emerged to cover these limitations and ll the gaps in the existing methodologies. Semantic web is one of these emerging technologies which is a ecting almost every single aspect of software engineering. The research on semantic web and ontology- based communication and knowledge sharing has shown promising results during the past few years. Recently there has been mushrooming of e orts on application of semantic web for software engineering. There are many systems built for semantic enabled software engineering but these systems are never been compared or cross evaluated. No metrics have been de ned for comparison and benchmarking of these systems. We have identi ed met- rics for comparison and benchmarking of these systems and categorized these metrics into relevant dimensions. The output of our research is an evaluation framework for comparative analysis and benchmarking of e orts in semantic web driven software engineering research. We then tested our framework on di erent semantic enabled software engineering systems, by evaluating them based on the metrics de ned in our framework. At the end, we have evaluated our work by doing a survey from software engineering and seman- tic web communities, by asking them to validate our identi ed metrics for comparison. The objective of our framework is to lead a way towards the improvement and maturity of the tools and technologies being developed in the area of semantic enabled software engineering. Our framework will help software industry, in checking the suitability of the available software devel- opment tools and research community, in evaluating their work and getting future directions for further research.