Pacific Asia Journal of the Association for Information Systems


Business intelligence (BI) has gained a strategic importance in today’s global competitive environment. However, high-quality BI is not easy to obtain on the Web due to information overload and difficulty to present complicated relationships among various types of business stakeholders. Unfortunately, existing BI tools lack the capability of analyzing and visualizing such relationships and research on BI systems is sparse. In this paper, we review the current market of BI tools and related research, describe an approach to support the development of tools that provide high-quality BI, and report the findings of a user evaluation study of the prototype developed based on the proposed approach. The approach combines information visualization and Web mining techniques with human knowledge to enable business analysts to analyze and visualize complicated business stakeholder relationships. Results of an experiment involving 62 subjects show that the prototype significantly outperformed a traditional method of BI analysis in terms of efficiency, quality of BI, and user satisfaction. The subjects provided favorable comments and expressed strong preferences toward the prototype in most applications. This research contributes to advancing BI research and to providing new empirical findings for BI systems evaluation.