Ontology Based Semantic Enhancement of Web Services Discovery

Abstract

Service Oriented Architecture (SOA) has become a major trend in the soft- ware industry for sometime now. Web services are a major enabling technol- ogy of the SOA. Today, most of the companies have exposed their services as web services. Since the number of web services available are very high, manual service discovery based on keywords has now become nearly impos- sible. So the use of automated service discover methods are critical. This is a hybrid service discovery method that uses both ontologies and Universal Description Discovery and Integration (UDDI). Our proposed approach can outperform the UDDI by using semantic capabilities. Unlike pure semantic service discovery methods, this system does not require prior knowledge on semantic technologies and has the ability to minimise the processing overhead of the requests by initially minimizing the result set using UDDI. Proposed architecture and implementation detail of a prototype of the proposed system is provided with this report. The evaluation by comparing with the UDDI has shown that our system has a signi cant improvement in precision and F1 scores, compared to UDDI registry. This approach stand out from other related hybrid methods, since it does not require any modi cation of the existing UDDI set-up. Companies can introduce our system to their existing UDDI set-up as a separate semantic layer