ANALYZING THE USEFULNESS OF THE USER’S BROWSER HISTORY FOR GENERATING QUERY SUGGESTIONS

Abstract

A very useful feature of search engines that helps users while they browse the internet, where, they very often, try to satisfy an information need, is query suggestion. This mechanism shows the user a list of possible queries from where he can choose and be able to perform a search easier and faster. In this paper we tried to assess the usefulness of a user’s recent web browsing history for generating new query suggestions. We performed a one month experiment in which we collected browsing history logs of several users and searched query terms submitted by those users to Google (using a Chrome plugin) and found that approximately 32% of the queries submitted can be predicted from the user’s browsing history.

Keywords

• query suggestion, browsing history, personalized suggestions.