Font Size: 

Despite the increasing amount of attention that Knowledge Management (KM) has received from academia, Universities have been slow to implement KM initiatives internally (Moss et al. 2007, Muthukumar and Hedberg 2005). Jones et al. (2006), however, report that such initiatives hold tremendous value for Higher Education Institutions, being the ultimate knowledge organisations (Jones et al. 2006). One such initiative is the development of a web-based Knowledge Directory to enable the location of expertise within a University, enhancing its ability to utilise its own knowledge assets as well as providing a means for external actors to explore the its knowledge network.

Dooley et al. (2002) report that 40%-50% of KM applications have focussed on the construction of expert profiles, emphasising it as a crucial success factor for KM systems. A key challenge, however, is ensuring that expert profiles are consistently updated. Even with functionality in place to enable employees to update their own profiles, the challenge of ensuring active user participation remains. Farell et al. (2007) report that, at IBM, one of the “pain points in the existing corporate profiling system has been the amount of time and complexity needed to update personal content”.

Social bookmarking systems, notably del.ico.us, have implemented the notion of tagging with much success and other Web 2.0 applications (Flickr and Technorati) have adopted and popularised the concept (Farell et al., 2007). Some efforts have also been made to implement the tagging of people, Farell et al. (2007) mention Tagalag (http://tagalag.com), 43people (http://43people.com) and Xing (http://xing.com) as notable examples. Likewise, IBM has implemented people-tagging functionality in their next-generation employee directory, Fringe. Although the functionality has had a positive impact on the maintenance of their expert profiles, they still report low adoption rates (346 out of 25 026 people were responsible for 85% of all tags created).

Academic environments have the distinct advantage that the knowledge of an expert is captured (at least partly) in academic publications or theses. In this study the possibility of constructing an expert profile based upon tags extracted from these knowledge artefacts is explored. If found to be feasible, the approach may enable the automation of expert profile construction. An experimental web-based knowledge directory was developed using PHP and MySQL. Ten experts with varying academic backgrounds were asked to participate in the study and their personal and contact information, as well as a short biography was recorded in the database. Experiments were conducted with various methods of text analysis to find a suitable approach to the construction of tag clouds based upon the combined text of an expert’s publications and theses. The chosen method consists of various phases and implements the Porter stemming algorithm. Subsequently, semi-structured interviews were conducted with the experts to determine whether the tag clouds were accurate representations of their expertise.