Abstract | ||
---|---|---|
Diverse strategically misbehaved entities have severely degraded the core-functions of trust-enabled interactional networks. At present, it is still a hard problem to identify them owing to the complexities of malicious behaviors, such as on-off attack, colluding attack, etc. In this paper, we propose a belief propagation-based algorithm Map-Trust to quantitatively and qualitatively infer entity's trustworthiness and untrustworthiness. Three primary contributions are included: (i) we define removal probability for each pair of interacted entities via pairwise feedback-ratings; (ii) we propose a novel cross-iteration fashion to infer trustworthiness and untrustworthiness values. The cross-iteration fashion not only declines time overhead compared to sequential iteration method, but it also supports a convenient manipulation, i.e. we can flexibly initiate group affinity; (iii) we launch extensive experiments using synthetic and real-world datasets to verify the efficiency of our proposed MapTrust. The experimental results show our proposed MapTrust dramatically outperforms Monte Carlo Markove Chain and Random algorithms against four representative attacks. |
Year | DOI | Venue |
---|---|---|
2018 | 10.1007/978-3-319-94289-6_25 | WEB SERVICES - ICWS 2018 |
Keywords | Field | DocType |
Trustworthiness propagation, Group assignment, Belief propagation, Trust-enabled interactional networks | Pairwise comparison,Data mining,Monte Carlo method,Inference,Iterative method,Trustworthiness,Computer science,Belief propagation | Conference |
Volume | ISSN | Citations |
10966 | 0302-9743 | 0 |
PageRank | References | Authors |
0.34 | 14 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Xinxin Fan | 1 | 16 | 5.10 |
Danyang He | 2 | 0 | 0.34 |
Jingping Bi | 3 | 70 | 18.36 |