Types of Document Search Tasks and Users' Cognitive Information Seeking Strategies

Abstract

For the researchers and learners, an unprecedented number of documents became available on the Internet and academic archives. Powerful search systems and sophisticated recommendation services are also available. Despite the IT assistance, finding the most useful information in daily knowledge works has become a cognitively demanding task more than ever due to the overwhelming number of documents. To improve the search systems with better human-computer cooperation, human information seeking strategies should be understood. This paper reports a study that identified the differences in the user search strategies with respect to two major search task types: open and purpose-driven exploring (OT) vs. closed and target-specified (CT) tasks. An observational experiment was conducted and the results were analyzed by mapping the user activities on a cognitive task-flow framework. The analysis comparing user activities in four search tasks revealed notable differences in their strategies to deal with the two task types. More frequent re-planning, especially goal reformulation, was observed for OT type tasks. The difference indicates that OT type tasks tended to trigger more knowledge-based behavior, while CT type tasks were performed relying more on rule-based behavior. These findings provide important insights for the design of search systems and user interfaces of knowledge-based systems.