Aiming at malicious fraud in the complicated network environment, a method framework combining service credibility evaluation and QoS-aware service composition optimization is proposed. Firstly, based on the historical behavior of Web services, Bayesian learning theory and evaluation information of historical users are used to evaluate the credibility of Web services from both objective and subjective aspects. Then, by using the service QoS attributes measured by credibility, a multi-objective optimization model is constructed, and an improved multi-objective grey wolf optimization(IMOGWO) algorithm is proposed for model solution. Finally, the effectiveness of the method framework for service composition optimization is verified by experimental data.