PRIVACY PROTECTION TECHNIQUES FOR WEB SEARCH PERSONALIZATION

Abstract

The search engine returns more relevant results when you utilize personalized web search (PWS). Users' reluctance to divulge personal information while searching is a key factor in PWS's low adoption rate, according to the study. Our research primarily focuses on how Personalized Web Services (PWS) applications that leverage hierarchical user accounts reveal consumer preferences. A novel PWS platform, UPS, is the topic of this essay. By utilizing adaptive query-based approaches, this system can generalize user profiles while still honoring their privacy choices. One predictive metric considers the value of personalization, while the other considers the potential danger of revealing your profile; our runtime extension seeks a happy medium between the two. As an example, we demonstrate how to increase the amount of computations at runtime using greedy dynamic programming and integer linear programming techniques. Additionally, you can assess the potential value of question customization using our online tool. Extensive experiments have confirmed our method's efficacy time and time again. According to the research, GreedyIL outperforms Greedy DP.