MEMORY: A Matrix-Based Efficient Semantic Web Service Discovery System

Abstract

With the advance of Semantic Web, the adoption of Semantic Web has been regarded as the most promising way to improve the recall rate and precision rate of service discovery. However, Semantic Web Service discovery (SWS) is essentially not used on a large scale in real business world due to its time consuming performance and weak support for the QoS-based discovery. In order to solve these problems, this paper presents a matrix-based efficient SWS discovery system, namely MEMORY. MEMORY does ontological pre-reasoning and holds the reasoning results in matrix forms in service publishing phase, so that it can transfer the load of semantic reasoning from service query to service publication and perform fast matching during service discovery. The experiments in the end are conducted to further demonstrate the feasibility of our proposed matching approach and its high efficiency.