Low-Cost Web Service Discovery Based on Distributed Decision Tree in P2P Environments

Peer to Peer (P2P) techniques were widely applied into Web service discovery fields, but most existing schemes only focused on matching the functional attributes of Web services and ignored the non-functional attributes, i.e. quality of services (QoS). In this paper, we present a QoS-aware Web service discovery approach. The approach maps each Web service into virtual space to distribute all QoS attributes among P2P nodes, and then designs a distributed decision tree to support range service query with QoS requirements. Finally, a load balance algorithm is used to reduce range query cost. Experimental results based on synthetic and real datasets show our proposed approach can perform effective service discovery with low cost in P2P environments.