Improving performance of semantic web services discovery with a hybrid matchmaker

The number of available Web services, nowadays, is growing rapidly due their potential in many fields. As a result, the discovery process becomes a challenging issue. Traditional syntactic keywords based discovery techniques are inefficient as they fail to recognize similarities between Web services capabilities. Thus, the most promising approach seems to be the introduction of semantic technologies. Existing Web services semantic-based discovery techniques rely only on subsumption relations between ontological concepts. On this basis, we present in this paper a novel discovery framework which performs a hybrid matchmaking that combines several similarity techniques to efficiently discover the most appropriate services to user requirements. Our approach involves formatting and extracting useful semantic information from user query expressed in natural language. We implemented our proposed approach in a prototype system called OWLS-LLCR. We also present some evaluation results of our system compared to the OWLS-MX matchmaker. Our experimental results validate the efficiency and the scalability of our approach.