Towards a Metrics Suite for Measuring Composite Service Granularity Level Appropriateness

Abstract

One of the prominent principles of designing services is the matter of how abstract services should be i.e. granularity. Since service-oriented analysis and design methods lack on providing a quantitative model for service granularity level evaluation, identification of optimally granular services is the key challenge in service-oriented solution development. This article through a systematic process proposes a model namely Weighted Granularity Level Appropriateness (WGLA) which leverages and consolidates four metrics to constitute quantitative basis for granularity appropriateness analysis. These metrics are, indeed, the four quantified attributes of service granularity including business value, reusability, context-independency, and complexity. Our preliminary controlled experiment confirms the correctness of the quantitative model. In fact, by adopting WGLA metric, service granularity appropriateness analysis could be conduct quantitatively that leads to realize an optimized service-oriented solution in terms of its granularity.