The Anatomy of a Click: Modeling User Behavior on Web Information Systems
Source: University of Texas
Providing relevant information to users with minimal effort on their part remains the holy grail of search science. Since the retrieval and presentation of search results as well as any dynamic search engine behavior depend on the estimated user intent, it is highly critical to accurately infer the user intent using all the available information on user session behavior at any stage. However, the necessity to minimize the user cognitive effort as well as intrinsic limits of web instrumentation require search engines to only have simplistic search user interface and logging mechanisms that can capture only a subset of user actions that are predictive of the search intent.