Services Discovery and Recommendation for Multi-datasource Access: Exploiting Semantic and Social Technologies

Abstract

The advent of Service Oriented Architectures (SoA) in the late 90s has significantly changed the development of enterprise systems. Web application development relying on selection and reuse of services, offered as third party software components, has been proposed as a new paradigm to effectively support creativity and productivity of developers. This development paradigm strongly requires advanced discovery and recommendation techniques, able to use and combine different types of information to suggest the most suitable data services for multi-datasource access. WSDL-based, semantic-enriched service matchmaking approaches have been initially proposed to enable service discovery and composition. Subsequently, approaches for web mashup, through RESTful services and Web APIs selection based on their lightweight descriptions, have emerged to meet requirements of agile development. Recently, in this context, service discovery and recommendation techniques are being empowered by considering factors related to the social web such as the existence of developers social networks and the possibility of evaluating the experience of web application developers. According to these premises, in this chapter, we present main features of a comprehensive data service selection framework, apt to provide advanced discovery and recommendation techniques. In the framework, an experience perspective will be considered, focused on social networks of developers, where social relationships represent explicit endorsements among developers concerning their skill in Web application development and votes on data services, assigned by developers, are used to estimate developers’ credibility according to a majority-based approach.