Estimating the complexity of software services using an entropy based metric

Abstract

Poor design of software services results in unnecessarily complex and inflexible SOA applications that are difficult to maintain and evolve. There is also some evidence that the quality of service-oriented applications degrades and as complexity increases with new service versions that include modifications to rectify problems and improve functionality. Design quality metrics play an important role in identifying software quality issues early in the software development lifecycle. The concept of software entropy has been used in literature to express decline in the quality, maintainability and understandability of software through its lifecycle. In this paper we propose a Service Entropy Metric (SEM) that estimates the complexity of service design based on the complexity of the XML message structures that form the basis for service interfaces. We illustrate the application of the SEM metric using the Open Travel Alliance specification and show that the complexity of the specification as measured by SEM increases over time as new versions of the specification are released.