Integrating behavioral trust and reputation in Web service compositions

Abstract

Algorithms for composing Web services traditionally utilize the functional and quality-of-service parameters of candidate services to decide which services to include in the composition. Users often have differing experiences with a Web Service. While trust in a Web Service is multi-faceted and consists of security and behavioral aspects, the focus in this thesis is on the latter. A formal model for trust in a Web Service is adopted, which meets several intuitions about trustworthy Web Services. In order to make the system robust an additional model to identify the reputation of users that provide feedback is introduced. Predictors of a positive experience with a Web Service and positive reputation of a user are hypothesized. A small pilot study is conducted to explore correlations between subjects' experiences with Web Services in a composition and the predictor values for those Web Services. Furthermore, the method to derive trust for compositions from trust models of individual services is presented. A novel framework, called Wisp, is presented and evaluated that utilizes the trust models and, in combination with any Web Service composition tool, chooses compositions to deploy that are deemed most trustworthy.