A B C D E F G H I J K L M N O P Q R S T U V W X Y Z All
Personalized Web Search (PWS) has demonstrated its usefulness in up the standard of assorted search services on the web. However, evidences show that user’s reluctance to reveal their private information throughout search has become a significant barrier for the wide proliferation of PWS. Personalized Search is one of the options available to users in order to sculpt search results returned to them based on their personal data provided to the search provider. We examine privacy protection in PWS applications that model user preferences as hierarchical user profiles. We tend to propose a PWS framework known as UPS which will adaptive generalize profiles by queries whereas respecting user such privacy requirements. Our runtime generalization aims at striking a balance between two predictive metrics that evaluate the utility of personalization and the privacy risk of exposing the generalized profile. We present two greedy algorithms, namely GreedyDP and GreedyIL, for runtime generalization. We also provide an online prediction mechanism for deciding whether personalizing a query is beneficial. Extensive experiments demonstrate the effectiveness of our framework. The experimental results also reveal that GreedyIL significantly outperforms GreedyDP in terms of efficiency.
