Kind of quantitative evaluation of user behaviour trust using AHP

LiQin Tian,Anjuan Qiao,Chuang Lin,Tieguo Ji

2007-01-01

Journal of Computational Information Systems

Abstract:In the daily life, decision about trust almost happen every day. In Internet such as search engine and Web mining, trust is also important and necessary, because the shared information environment was open, dynamic, uncertain and heterogeneous, it is very difficult to completely acquire all security information, which results in increasing unreliability and security threat in Internet. One of the most important methods to solve this problem is to establish trust between these entities named user trust. User trust can avoid or reduce interaction with the malice user to improve security. User trust includes two aspects, one is user identification trust and the other is user behaviour trust. The paper discusses a kind of quantitative evaluation of user behaviour trust using Analytic Hierarchy Process (AHP) theory. We put forward the hierarchical structure model used to evaluate user behaviour trust and we use an idea of 'divide and rule' based on hierarchy and decomposition to decompose complicated user behaviour trust into small specific behaviour trust attributes and trust evidence, then we evaluate the weight of each behaviour trust evidences and behaviour trust attributes based on AHP and finally we use evidence matrix and weight matrix to compute user behaviour trust.