Semantic discovery of web services using statistical methods and measures of semantic relatedness

Discovering a service over the web that meets the desired functionalities is still one of the most challenging tasks in the area of Service Oriented Computing. Lack of semantic information in the web services profiles poses a restriction in the automated discovery of services. Irrelevant and huge number of services returned by the UDDI and lack of standard mechanisms are the main problems faced by the users today during service discovery. In this paper, we propose a Web service discovery approach independent of the description model that tries to manage with the heterogeneity found in semantic service description frameworks. This proposed approach uses the principals from text mining, measures of semantic relatedness and information retrieval where the semantic information of the services is integrated with the syntactic service profiles to give hybrid service vectors. Empirical evaluation of the proposed approach implemented on OWL-X services has been presented to show the feasibility of the approach. Experimental results have shown that the proposed approach is able to discover better semantic relationship between services, therefore, more relevant results are ensured during discovery.