Semantic Web Service Discovery Using Natural Language Processing Techniques

Semantic Web Service Discovery Using Natural Language Processing Techniques

Abstract— the proliferation of available Web services, new approaches are needed to search for the most appropriate services matching user requirements. Existing UDDI based discovery techniques fail to recognize similarities and differences between Web service capabilities and thus greatly limit the scope of service discovery. Hence, in this paper we present a framework for discovering the vast amount of Web services present< Final Year Project > in Internet based on a novel semantic approach to automatically discover the most appropriate services to user requirements. Our framework is based on a natural language query to facilitate usage of our implemented discovery framework. For this purpose, we first apply some computational linguistics techniques such as part-of-speech tagging and word sense disambiguation to format and extract useful semantic information from user query. Then, we propose an innovative semantic matchmaking technique based on significant concepts properties in the referenced domain ontology and we show that it improves discovery efficiency compared to a similar related work.