Abstract
In this study, a proposed document (knowledge) management system based on a text-mining technique will be tested in an experiment manner. Two independent variables, task complexity and system functionality, are chosen in this experiment.Thus, the main goal of this study will be: 1) to understand how different level of task complexity would affect the system performance; 2) to understand how different system functionality would affect the system performance; 3) to understand how the interaction effect of task complexity and system functionality on the system performance.The physical configuration of proposed system will be measured in comparing with another classification algorithm in efficiency and effectiveness manners.Experimental and analytical results show that system functionality significantly influence system performance perceived by users. Although another effect of task complexity on system performance is not significant, we still discuss the research limitations and provide some suggestions to overcome them for future experiments and researchers.