Enhanced Utility in Search Log Publishing of Infrequent Items

Abstract

Search engines are being widely used by the web users. The search engine companies are concerned to produce best search results. Search logs are the records which records the interactions between the user and the search engine. Various search patterns, user's behaviors can be analyzed from these logs, which will help to enhance the search results. Publishing these search logs to third party for analysis is a privacy issue. Zealous algorithm of filtering the frequent search items in the search log looses its utility in the course of providing privacy. The proposed confess algorithm extends the work by qualifying the infrequent search items in the log which tends to increase the utility of the search log by preserving the privacy. Confess algorithm involves qualifying the infrequent keywords, URL clicks in the search log and publishing it along with the frequent items.