Wikimedia Research/Showcase/Archive/2015/03

March 2015

March 25, 2015 Video: YouTube

User Session Identification Based on Strong Regularities in Inter-activity Time

By Aaron Halfaker

Session identification is a common strategy used to develop metrics for web analytics and behavioral analyses of user-facing systems. Past work has argued that session identification strategies based on an inactivity threshold is inherently arbitrary or advocated that thresholds be set at about 30 minutes. In this work, we demonstrate a strong regularity in the temporal rhythms of user initiated events across several different domains of online activity (incl. video gaming, search, page views and volunteer contributions). We describe a methodology for identifying clusters of user activity and argue that regularity with which these activity clusters appear implies a good rule-of-thumb inactivity threshold of about 1 hour. We conclude with implications that these temporal rhythms may have for system design based on our observations and theories of goal-directed human activity.

Mining Missing Hyperlinks from Human Navigation Traces

By Bob West

Wikipedia relies crucially on the links between articles, but important links are often missing. In most prior work, the problem of detecting missing links is addressed by constructing a model of the existing link structure and then predicting the missing links based on this model. In this work we propose a novel method that does not rely on such a model of the static structure of existing links, but rather starts from data capturing how these links are used by people. The approach is guided by the intuition that the ultimate purpose of hyperlinks is to aid navigation, so we argue that the objective should be to suggest links that are likely to be clicked by users. In a nutshell, our algorithm suggests an as yet non-existent link from S to T for addition if users who open S are much more likely than random to later also open T. We show that this simple algorithm yields good link suggestions when run on data from the human-computation game Wikispeedia.net. Finally, we show preliminary results that show the method also works "in the wild", i.e., on navigation data mined directly from Wikipedia's server logs.