Evaluating online trust using machine learning methods

Trust plays an important role in e-commerce, P2P networks, and information filtering. Current challenges in trust evaluations include: (1) fnding trustworthy recommenders, (2) aggregating heterogeneous trust recommendations of different trust standards based on correlated observations and different evaluation processes, and (3) managing efficiently large trust systems where users may be sparsely connected and have multiple local reputations. The purpose of this dissertation is to provide solutions to these three challenges by applying ordered depth-first search, neural network, and hidden Markov model techniques. It designs an opinion filtered recommendation trust model to derive personal trust from heterogeneous recommendations; develops a reputation model to evaluate recommenders\u27 trustworthiness and expertise; and constructs a distributed trust system and a global reputation model to achieve efficient trust computing and management. The experimental results show that the proposed three trust models are reliable. The contributions lie in: (1) novel application of neural networks in recommendation trust evaluation and distributed trust management; (2) adaptivity of the proposed neural network-based trust models to accommodate dynamic and multifacet properties of trust; (3) robustness of the neural network-based trust models to the noise in training data, such as deceptive recommendations; (4) efficiency and parallelism of computation and load balance in distributed trust evaluations; and (5) novel application of Hidden Markov Models in recommenders\u27 reputation evaluation