Measuring Service Cohesion Using Latent Semantic Indexing
Ali Kazemi, Ali Rostampour, Fereidoon Shams, Pooyan Jamshidi, Ali Nasirzadeh Azizkandi · International Conference on Internet and Web Applications and Services · 2011
As low coupling, high cohesion is a service-oriented design and development principle that should be kept in mind during all stages. High cohesion increases the clarity and ease of comprehension of the design that simplifies maintenance and achieves service granularity at a fairly reasonable level. However, unlike coupling that only measures the degree of structural and behavioral dependency to the other services, cohesion metrics need to evaluate the degree of semantic relationships between operations within a service in order to measure functional relatedness. Latent semantic indexing (LSI) is one of the techniques in the field of information retrieval which is widely used to measure the degree of semantic relatedness between a document and a given query and also used to measure the cohesion of a text. In this paper, we propose an approach to automatically measure the strength of conceptual cohesion of a service based on LSI technique. Finally, it has been evaluated theoretically based on a set of cohesion principles. Keywords-Servivce cohesion; Latent semantic indexing; Software metric.