点击次数：

论文类型：期刊论文

发表时间：2012-01-01

发表刊物：JOURNAL OF INTERNET TECHNOLOGY

收录刊物：SCIE、EI、Scopus

卷号：13

期号：1

页面范围：159-172

ISSN号：1607-9264

关键字：Multi-agent system; Fuzzy set theory; Trust; Lattice measure of similarity; Fuzzy inference

摘要：Trust mechanism is the key foundation to ensure the security in multi-agent environments. In view of the shortcomings in the existing models, we develop a valuable evaluation mechanism using fuzzy set theory to describe its trust models in this paper. A multiple-attribute fuzzy decision is introduced so that the trust evaluation is more consistent with the thinking custom in the human society. In our trust recommendation, we present similarity techniques based on the coefficients of rating-attribute weight so that the trust values are more accordant with subjectivity, and introduce a lattice measure of similarity that can effectively diminish the influence from some malicious recommendations. Meanwhile, our model improves the derivation rules of trust recommendation relationships by multiple multivariable fuzzy reasoning mode and fuzzy uncertainty reasoning theory. Subsequent experimental results show that our model can inhibit malicious transactions more effectively than Eigen Trust, and thus give a significant way for studying the security mechanism in multi-agent environments.