The cognitive structure of user sense making (Keynote talk)
Digital libraries and visual interfaces are often designed to be a part of large sensemaking tasks, such as understanding a medical problem or developing a scientific model. We have been studying sensemaking tasks in two thrusts. The first thrust involves experiments with users performing subparts of the overall sensemaking task. For instance, we have developed cognitive models of finding information with an advanced visual tree browser and exploratory data analysis with an advanced visual table browser. In our second thrust we find experts who excel in sensemaking and use Cognitive Task Analysis to yield information about their knowledge, thought processes, and goal structures. Our recent focus has been on intelligence analysts who make predictions from crushing amounts of semantically rich data. The study of expert sensemakers provides a window to the cognitive structure of a complex sensemaking task that we believe has much in common with other sensemaking tasks. We found that the overall process of analysis is an opportunistic mix of two major loops of activities: (1) a foraging loop that involves processes aimed at seeking information, searching and filtering it, and reading and extracting information, and (2) a sense making loop that involves iterative development of a mental model (a conceptualization) that best fits the evidence. Information processing can be driven by bottom-up processes (from data to theory) or top-down (from theory to data). In my talk I will review our initial understanding of these processes, discuss leverage points for sensemaking technologies, and suggest metrics for the evaluation of sensemaking technologies.
Pirolli, P. L. The cognitive structure of user sense making. (Keynote presentation) Seventh International Workshop on Audio-Visual Content and Information Visualization in Digital Libraries (AVIVDiLib'05); 2005 May 4-6; Cortona; Italy.