Web service selection based on similarity evaluation
Non-functional property is of paramount importance for Web services to function. However, service selection with consideration of Quality of Service is a challenge. This paper presents an approach to service selection by measuring the similarity between the web service publication and the service request. We identify five tendencies of the nonfunctional dimensions in the service requests, and present a flexible matchmaking approach which enables users to preset the negotiable preferences. Unlike other service selection approaches, we obtain the relative distance to compute the similarity score for generating the ranking list. Specifically, we introduce a novel method for evaluating the similarity between two categorical values based on the semantics and set theory.