Resisting Collusion by Game in Culture Web

• Main Authors: Bao Yu, Cao Tianjie, Zeng Guosun
• Format: Conference Proceeding
• Language: English
• Subjects: Algorithm design and analysis; collusion recognition; Computational modeling; Cultural differences; culture web; Game theory; Games; Peer to peer computing; Reliability; reputation system; trust system
• ISSN: 1524-4547; 2641-8169
• DOI: 10.1109/WETICE.2011.45

State of the culture web allow the users to aggregate information from many culture heritages' sources or buy the foreign culture products. In these websites, reputation systems, a proved useful mechanism, is introduced to help people to decide which information deserved to trust. To maintain the overall credibility of the information, users are confronted with assessing the reliability of information from different sources. This paper presents anti-collusion algorithm, a collusive group recognition method based on game theory. In reputation systems, users forming collusive group tend to increase theirs own reputations by abusing evaluation of other users' behaviors. Anti-collusion algorithm makes a good use of the game between the members of collusive group, who have to form a special pattern of profit distribution not only on income of reputation but also on distribution of profit if they want to get the benefits from collusion. In order to promote capabilities of the system, anti-collusion algorithm will be adopted to distinguish these groups in terms of the special pattern.