Research on interest domain partition based P2P intuitionistic fuzzy trust model

Trust in Peer to Peer network has the fuzzy feature, which cannot be represented by most of current distinct mathematic theory based trust models. To solve the problem, an interest domain partition based P2P intuitionistic fuzzy trust model IFTrust is proposed, which uses interest domains to represent the interest biases of different peers, and uses trust factors set as well as trust levels set to describe the service quality of different peers. The integrate information of a peer´s trust is represented by the membership, nonmembership as well as hesitancy degree to the upper intuitionistic fuzzy sets. Besides, when calculating the domain service reputation of a peer, IFTrust uses domain recommendation similarity to weight the domain recommendation vectors provided by different peers. Simulation results show that IFTrust can represent the fuzzy and domain features of trust more accurately, and the successful download ratio as well as security level of IFTrust is better than that of traditional models.