Article ID: | iaor2008643 |
Country: | Netherlands |
Volume: | 42 |
Issue: | 3 |
Start Page Number: | 1539 |
End Page Number: | 1557 |
Publication Date: | Dec 2006 |
Journal: | Decision Support Systems |
Authors: | Chen Wun-Hwa, Wu Soushan, Chen Hsinchun, Huang Zan, Xu Jennifer J., Guo Fei |
Keywords: | artificial intelligence: decision support |
Expertise management systems are being widely adopted in organizations to manage tacit knowledge. These systems have successfully applied many information technologies developed for document management to support collection, processing, and distribution of expertise information. In this paper, we report a study on the potential of applying visualization techniques to support more effective and efficient exploration of the expertise information space. We implemented two widely applied dimensionality reduction visualization techniques, the self-organizing map (SOM) and multidimensional scaling (MDS), to generate compact but distorted (due to the dimensionality reduction) map visualizations for an expertise data set. We tested cognitive fit theory in our context by comparing the SOM and MDS displays with a standard table display for five tasks selected from a low-level, domain-independent visual task taxonomy. The experimental results based on a survey data set of research expertise of the business school professors suggested that using both SOM and MDS visualizations is more efficient than using the table display for the associate, compare, distinguish, and cluster tasks, but not the rank task. Users generally achieved comparable effectiveness for all tasks using the tabular and map displays in our study.