Usage data in web search: benefits and limitations

摘要: Web Search, which takes its root in the mature field of information retrieval, evolved tremendously over last 20 years. The encountered first revolution when it started to deal with huge amounts pages. Then, a major step was accomplished engines consider structure graph and link analysis became differentiator both crawling ranking. Finally, more discrete, but not less critical step, made search monitor mine numerous (mostly implicit) signals provided by users while interacting engine. We focus here on this third "revolution" large scale usage data. detail different shapes takes, illustrating benefits through review some winning features that could have been possible without it. also discuss limitations how cases even conflicts natural users' aspirations such as personalization privacy. conclude discussing these can be circumvented using adequate aggregation principles create "ad hoc" crowds.