USABILITY AND PROBABILISTIC MODELING FORINFORMATION SHARING IN DISTRIBUTED COMMUNITIES

摘要

This paper describes work done in developing realistic simulations of information sharingin distributed expert communities. Instead of a rule-based rational behavior approach, wetake a probabilistic modeling approach for studying group performance and timeconstrained,socially-mediated tasks. Four simulation modules for advising, sharing,solving and learning had been proposed and are in development. This work presentsresearch done using an online blog to gather supporting usability data for the secondmodule. This is put in context of earlier work suggesting use of the above simulationmodules. We present data from two studies suggesting the importance of socialinteraction in distributed expert communities. The results presents make a strong case forusing the proposed simulation modules as a design and usability tool.