A Web Service Discovery Scheme Based on Structural and Semantic Similarity

With the increasing adoption of Web Services and service-oriented computing paradigm, matchmaking of web services with the request has become a significant task. This warrants the need to establish an effective and reliable Web Service discovery. Here reducing the service discovery time and increasing the quality of discovery are key issues. This paper proposes a new semantic Web Service discovery scheme where the similarity between the query and service is decided using the WSDL specification and ontology, and the improved Hungarian algorithm is applied to quickly find the maximum match. The proposed approach utilizes the structure of datatype and operation, and natural language description used for information retrieval. Computer simulation reveals that the proposed scheme substantially increases the quality of service discovery compared to the existing schemes in terms of precision, recall rate, and F-measure. Moreover, the proposed scheme allows consistently smaller discovery time, while the improvement gets more significant as the number of compared parameters increases.