Understanding Search Behavior Bias in Wikipedia.

As Wikipedia is one of the most visited sites on the Web, providing content to millions of readers daily from across the globe in more than 300 actively edited languages, it is relevant to understand the search behavior of its users. Extensive research over recent years has also highlighted the importance of studying the existence and potential impact of biases on the search process. Thus, based on Wikipedia’s server logs we generate data sets containing millions of search sessions to characterize its website search and compare it to generic web search. We segment the studied quantities by the type of client being used and analyze whether there is a relation between the time a user spends on a page and its length, as well as between the time it takes a user to click on a result and the position of the selected page in the ranking. We rediscover known results in the context of web search that also apply to website search, along with new results in the context of website search.