A Guided Mashup Framework for Rapid Software Analysis Service Composition

Historical data about software projects is stored in repositories such as version control, bug tracking and mailing lists. Analyzing such data is vital to discover unthought-of-yet-interesting insights of a software project. Even though a wide range of software analysis techniques are already available, integration of such analyses is yet to be systematically addressed. Inspired from the recently introduced concept of Software as a Service, our research group investigated the concept of Software Analysis as a Service (SOFAS), a distributed and collaborative software analysis platform. SOFAS allows software analyses to be accessed, composed into workflows, and executed over the Internet. However, traditional service composition is a complex, time consuming and error-prone process, which requires experts in both composition languages and existing standards. In this paper, we propose a mashup platform to address the problem of software analysis composition in a light-weight, programming-free process-centric way. Our proposed mashup platform provides design-time guidance to the users throughout the mashup design by integrating a continuous feedback mechanism. It requires exploiting semantic web technologies and Software Engineering Ontologies (SEON).