Understanding User Search Behavior Across Varying Cognitive Levels

The ubiquitous accessibility of the world-wide web has led people to increasingly use web search to learn or acquire new knowledge. Recent research efforts have targeted the optimization of web search to satisfy learning related needs. However, there is little known about how one's search interactions differ across varying cognitive levels that correspond to one's learning. In this paper, we address this knowledge gap by investigating how the search interactions of 150 users vary across 6 search tasks corresponding to distinct cognitive levels. We also analyze how users' knowledge gain varies across the cognitive levels. Our findings suggest that the cognitive learning level of a user in a search session has a significant impact on the user's search behavior and knowledge gain. Estimating the cognitive level of users during their interactions with search systems will allow us to construct and improve learning experiences for the users. For example, learners can be served content that corresponds to their current cognitive level within their learning process.