An Efficient Service Discovery Method and its Application

Abstract

To discover services efficiently has been regarded as one of important issues in the area of Service Oriented Computing (SOC). This article carries out a survey on the issue and points out the problems for the current semantic-based service discovery approaches. After that, an information model for registered services is proposed. Based on the model, it brings forward a two-phase semantic-based service discovery method which supports both the operation matchmaking and operation-composition matchmaking. The authors import the bipartite graph matching to improve the efficiency of matchmaking. An implementation of the proposed method is presented. A series of experiments show that the method gains better performance on both discovery recall rate and precision than a traditional matchmaker and it also scales well with the number of services being accessed.