Improve the Accuracy of Dirichlet Reputation System for Web Services

With existing a huge amount of services these days, too many web services have been developed with the same functionality. The problem arises that consumers of those services (usually designers of service-oriented applications) cannot blindly distinguish good services from bad ones. Therefore, finding effective and objective reputation for web services is important research topic. In this paper, Dirichlet reputation model has been improved for calculating reputation score of web services. The model has been applied using DS-DREAM dataset. For comparison purpose, the proposed system has been validated using QWS dataset and considered it as real score.