Semester Thesis for the: Information Technology and Electrical Engineering MSc
System Security Group, Department of Computer Science, ETH Zurich
Supervisors: Arthur Gervais, Prof. Dr. Srdjan Capkun
May 25, 2015
Effective web search personalization can be of paramount importance for user satisfaction and eventually web search engine's revenue. However, tracking users to provide personalization intrudes their privacy, especially when users do not voluntarily consent to it and provides biased web search results. In this work we develop a Chrome browser extension in order to examine how short term web search behavior affects web search results and we define metrics to quantify personalization. Overall, we conclude that personalization due to web search behavior is not immediate and bold, while interacting with web search results does not contribute further to personalization.