Poshida II, a Multi Group Distributed Peer to Peer Protocol for Private Web Search

Web Search Engines allow the user to retrieve data from the Internet using search query. WSEs maintain search query log to improve the result quality by retrieving personalized search result. WSE are known to sell query log related data to the marketing organizations for boosting revenues, which poses a threat to user privacy. Private web search is a concept in which the user gets data from the Internet through WSE while masking his identity. This work proposes Poshida II, a multi group distributed peer to peer protocol that preserves users privacy. The protocol is simulated using the AOL query log. The experiments are performed while varying number of users in multiple groups. The results are evaluated by measuring profile exposure level. Results show that Poshida II manage to hide user profile by 35% at first degree, 70% at second degree, 83% at third degree and 85% at fourth degree from WSE.