Personalized web service ranking via user group combining association rule

Abstract

Web service plays an important role in implementing Service Oriented Architecture (SOA) for achieving dynamic business process. With the increased number of web services advertised in public repository, it is becoming vital to provide an efficient web service discovery and selection mechanism with respect to a user's requirement. Considerable efforts have been made to solve this problem among which semantic based web service discovery has been attained much importance by researchers in academic and industry community. However, there is a challenge in the semantic based web service discovery process, that is, among the retrieved set of semantically equivalent web service candidates, how to discern which one is the best? In this paper, inspired by collaborative filtering idea, a web service ranking framework is proposed in which a set of users with similar interest will be firstly identified. Afterwards, association rules will be found out by analyzing all web service composition transactions related to that set of users. By combining user group and association rule mined from that group, a personalized web service ranking mechanism is achieved and the experiment shows the promising result. © 2009 IEEE.