Secured Personalized Web Search usingOntology and Hashing Technique

Abstract

Personalized Web Search is a best way to improve search quality by customizing search results for the users with individual information goals. This is achieved through implicitly collection of user profiles, browsing history, clicked through data, bookmarks, location of the user. However, users are not willing to expose their private preference information to search engines. Many research are undergoing in improving Personalized Web Search (PWS), this system added another dimension that is Privacy Protection in PWS. This system has a framework by name personalized privacy preserving search framework, which is very helpful framework where eavesdroppers are not able to get any details of the users since privacy measures are taking place at client side. To achieve privacy protection, this system provides metadata and user?s query encryption which is achieved using MD-5 hashing technique. In the beginning it will look like a keyword search but there is a behavioral observing system called the Spy-NB, which monitors the users? behavior and which will provide the results sets according to the users? interest in efficient way. To improve the personalized search, this system uses Taxonomy (ontology) concepts, RSMV, GreedyIL & GreedyDP algorithm. This system may be applied in real world applications like Google and Yahoo search engine, which allows users to describe their interests explicitly by selecting from pre-defined preference option, so that the results that match are re-rank according to the user?s interests and also gives users the option to save web sites they like and block those they dislike.