A Framework for Recommendation and Trust Assessment in SNS

With the development of the network technology, social network services (SNS), which are based on social network theory have gained increasing media, government and popular attention than ever before. In a social networking site, recommendation and trust assessment is of great importance. However, applications are complicated and changeable, so the fixed assessment criteria which are dedicated to one application, may not fit another. After analysis of specific criteria and in accordance with their properties, we divide the criteria into several categories of similar, monotonic, anti-monotonic, convertible and inconvertible, meantime, propose methods to deal with these criteria. We define a framework, and experimentally evaluate the effectiveness of our method.