Reputation-aware Service Selection based on QoS Similarity

For the up-and-coming computing models like as cloud computing, service is the standard package for meeting all kinds of consumers' requirements. Web Services are the concrete implement of the service. When users request and consume Web Services, services' reputations will play a vital role in users' selection. A gradually adjusting reputation evaluation method of Web Services is proposed based on eliminating the collusive behaviors of consumers step by step, and a reputation-aware model for service selection is designed. In order to adjust reputations, QoS similarity is computed firstly according to the differences between advertised QoS from service providers and delivered QoS from service consumers' evaluation, next, current reputation is attained; then the consumers are sorted based on reputation using clustering algorithm and the potential collusive consumers are mined using association rules algorithm; finally, the updated reputation is recalculated and saved in the reputation center included in the model. The experimental results show that the model can identify the malicious consumers and improve the exact rate of reputation evaluation and success rate of service selection.