Anti-Index: Against Privacy Mining via Search Engines

Xiaofeng Meng,Zhongyuan Wang,Jing Ai

Abstract:With the flourishing of Web 2.0, Internet has become the largest information depositary, which contains huge personal information of Web users. The information related to a specific person is usually scattered on various pages in different websites. However, today’s Internet has been highly crawled and indexed by search engines. The malicious attacker may collect a specific person’s information via search engines and obtain some privacy-sensitive information. Therefore, we observe a new type of privacy problem on the Internet: Privacy Mining via Search Engines. Our experiment shows this problem is serious yet easily ignored; it is a potential threat to Web users. We present a privacy mining model for describing the process of privacy leakage via search engines. To prevent this kind of privacy attack, we propose a method called Anti-Index and extend robots.txt to ERobots.txt with the fine-grained access control policies from the perspective of website constructors. In addition, we suggest a new service: automatically detecting potential dangers of privacy leakage for Web users. We also discuss several challenging problems about the privacy mining via search engines.

Computer Science,Engineering