A Probabilistic Approach for Web Service Discovery

Chune Li,Richong Zhang,Jinpeng Huai,Xiaohui Guo,Hailong Sun

DOI: https://doi.org/10.1109/SCC.2013.107

2013-01-01

Abstract:Web service discovery is a vital problem in service computing with the increasing number of services. Existing service discovery approaches merely focus on WSDLbased keyword search, semantic matching based on domain knowledge or ontologies, or QoS-based recommendations. The keyword search omits the underlying correlations and semantic knowledge or QoS information is not always available. In this paper, we propose a probabilistic service discovery approach to help web service users to retrieve related services and to improve the search performance. Specifically, we apply a probabilistic model to characterize the latten topics between services and queries, and then propose a matching method based on the topic relevance. Experiments on services from a real service repository confirm the feasibility and efficiency of this proposed method.