Web-based social networks have become one of the most popular applications on the Internet in recent years. However, most of the social networks rely on some simplistic trust models to manage trust information of the users, which can cause problems ranging from unsatisfied user experience to exposure to malicious users. This thesis proposes a gravity-based trust model to enhance the aggregation of personal trust information into a subjective reputation system. This new model maps all users on the social network into n-dimensional Euclidean spaces based on their direct trust information, and creates a trust social neighborhood for each user. The reputation of a target user is determined by applying gravity model to the information from both target's and observer's trust social neighborhood. A prototype of this trust model is implemented in order to evaluate the effects of varying different parameters of the gravity-based trust model.

No description available.