Comparison of two graph analysis techniques applied to semantic maps used to determine cohesiveness of object-oriented code

Software metrics refers to those quantitative measures that enable computer scientists to gain insight into the efficacy of the software process and the projects that are conducted using the process as framework. Focus on improving the software process development has spurred the need to identify measures, or metrics needed to manage the process. In fact, substantial research is being done in the field of software development to study the concept of software measurement techniques. It is important to measure software metrics so as to enable software managers and practitioners to improve the software process; assist in the planning, tracking, and control of a software project; and assess the quality of the software product. From an object-oriented viewpoint, semantic metrics are the metrics derived from the semantic aspects of the OO code.; The goal of the thesis is to compare the performance of the Bubble-clustering algorithm and Dijkstra's shortest path algorithm in generating object oriented (OO) software metrics. In this thesis, two different graph techniques, viz., Dijkstra's shortest path algorithm and a one-dimensional Bubble-clustering algorithm have been implemented. Furthermore, measures and metrics have been generated using these two graph approaches. These metrics have been applied on a collection of concepts and the results obtained thereby are compared against the results of metrics that were generated by human experts. A brief analysis of how the expert data correlates with the metrics produced by the two algorithms provides an insight into the performance of the algorithms in generating object-oriented software metrics.