Patterns of reading and organizing information in document triage

People engaged in knowledge work must often rapidly identify valuable material from within large sets of potentially relevant documents. Document triage is a type of sensemaking task that involves skimming documents to get a sense of their content, evaluating documents to assess their worth in the context of the current activity, and organizing documents to prepare for their subsequent use and more in-depth reading. We have performed a study of document triage by collecting multiple forms of qualitative and quantitative data to characterize how 24 subjects read about a new topic and assessed and organized a set of 40 relevant Web documents. Our results indicate that there are multiple strategies for document triage, each involving different styles of reading, interacting, and organizing. Common strategies include: 1) focused reading early in the task, relegating the organizing until later in the process; 2) skimming performed in tandem with organizing, which relies on gaining an incremental understanding of the topic; and 3) metadata-based organizing, a strategy that stresses working with document surrogates to minimize the time spent reading. The findings suggest ways applications may better support the intertwined nature of the browsing, reading, and organizing activities in document triage.