An Automatic Classifcation Approach to Idea Organization in Group Support Systems

With the rapid advance of information technologies, human beings increasingly rely on computers to accumulate, process, and make use of data. Knowledge discovery techniques have been proposed to automatically search large volumes of data for patterns. Group Support Systems (GSS) play an important role in streamlining group activities and improving group outcomes. Various attempts have been made to help automate several tasks in group activities under GSS environment. One of the approaches is to apply automatic approach to idea organization task in GSS. This research designed and tested a system using singular value decomposition (SVD) techniques to automate the idea organization task in GSS. Specifically, this research was conducted to examine how the idea organization task, typically regarded as the most labor-intensive and cognitively demanding in group problems solving, can be automated using a system enabled by singular value decomposition techniques. For the purpose of evaluation, we compared the performance of our automated approach using SVD algorithm against that of human subjects. Two separate experiments were conducted to evaluate the performance of the automatic approach on two essential components of an idea organization; namely, generation of topic categories and placement of relevant comments into their respective categories. The general conclusion that can be made from this research is that idea organization in group problem solving can be facilitated both efficiently and effectively with the SVDenabled system that can automatically generate topic categories and place relevant