Combining Global and Local Trust for Service Recommendation

Recommending trusted services to users is of paramount value in service-oriented environments. Reputation has been widely used to measure the trustworthiness of services, and various reputation models for service recommendation have been proposed. Reputation is basically a global trust score obtained by aggregating trust from a community of users, which could be conflicting with an individual´s personal opinion on the service. Evaluating a service´s trustworthiness locally based on the evaluating user´s own or his/her friends´ experiences is sometimes more accurate. However, local trust assessment may fail to work when no trust path from an evaluating user to a target service exists. This paper proposes a hybrid trust-aware service recommendation method for service-oriented environment with social networks via combining global trust and local trust evaluation. A global trust metric and a local trust metric are firstly presented, and then a strategy for combining them to predict the final trust of service is proposed. To evaluate the proposed method´s performance, we conducted several simulations based on a synthesized dataset. The simulation results show that our proposed method outperforms the other methods in service recommendation.