Supporting representation construction in sensemaking.

This study investigates how people construct representations in sensemaking tasks and explores the design space of systems supporting representation construction. Russell et al (1993) defined sensemaking as a process that searches for appropriate representation to answer task-specific questions. Their approach brought representation into the center of sensemaking research. My research follows this representational approach, narrowing it down to the construction of external and explicit representations in the sensemaking process because these representations provide a medium for people and computer systems to access and interact with the sensemaking process. The research approach is to interlace theoretical analyses and behavioral studies with system design and development in a spiraling process towards greater understanding of and effective support for sensemaking activities. The research starts with the examination of existing sensemaking theories, which lead to the development of a pilot sensemaking supporting system. Using this pilot system, a behavioral study was conducted to investigate representation construction in the early stage of sensemaking. Design implications were drawn from this experiment. Then a preliminary design space of systems supporting structure finding and exploration on the web was laid out. A system framework was proposed and a prototype was implemented. The system prototype provides a platform for further explorations in the design space and for further behavioral studies. This study finds that people not only get ideas for structuring representation from their existing knowledge, but also from the outside information world. Representation construction and information seeking are closely coupled in the process. The findings suggest important revisions of previous sensemaking theories and new opportunities for system design. A system framework is then proposed to help people locate and explore websites with high probabilities of containing useful structures. Many theoretic and design problems are identified for future exploration.Ph.D.Applied SciencesCommunication and the ArtsComputer scienceInformation scienceUniversity of Michigan, Horace H. Rackham School of Graduate Studieshttp://deepblue.lib.umich.edu/bitstream/2027.42/126268/2/3238064.pd