Skip to Main Content
A large number of works have been made to apply P2P techniques into services discovery. However, most of them only focus on matching the functional properties and ignore the non-functional properties. With a wide variety of services are offered that can satisfy the functional requirements, a policy is needed to support users to obtain the rank of the located services according to their nonfunctional preferences, and then select the appropriate services. Actually, the Web services which fulfill the functional requirement and meet or exceed the non-functional requirement are the candidates. In this paper, we present a structure P2P framework based on distributed decision tree(DDT) to support range queries in multiple QoS attribute space, and propose the algorithm to effectively distribute Web services among nodes to balance workload. Experiments with synthetic and real datasets show the effectiveness of the proposed scheme.
Date of Conference: 18-20 Oct. 2009