Privacy preserving personalization in probabilistic search

Abstract

Nowadays users want to retrieve accurate and relevant information as fast as possible by preserving their privacy. The main challenge in ensuring privacy is to provide accurate results to user as well as to hide relevant information from external entities. To provide the security in information retrieval system the relevant information must be protected from external entities. The proposed mechanism named Privacy Preserving Probabilistic Search, provides the necessary privacy for a particular personalised search. The users profile helps to get essential input for personalized search. The main motivation for ontological query expansion is, to improve results by including terms that would lead to retrieving more relevant documents. In the proposed method, both the keyword and the entities interest are made hidden. An encrypted data scheme is used for uploading and retrieval of documents without loss of privacy. The probabilistic search ensure excellent result with high recall.