Expertise location: can text mining help?

Abstract

Tell me where you surf, and I will tell you who you are In Ackerman's view [l] Computer Supported Knowledge Management falls into two categories: one to support knowledge externalization by creating knowledge repositories depending on well established computational techniques and the other sharing which is the new generation of KM work. Part of the sharing domain is which entails defining what is an expert? Measuring the level and determining with a degree of assurance that this expert would be willing to collaborate. On the other hand, Natural language text analysis was developed with the goal of rapid and efficient navigation through large volumes of textual information helping numerous professions that require such activities, and which benefit from automation where possible. There is a group of technologies under the heading that deal with this automation. Under this group of technologies comes: extraction of significant data from text, a degree of categorization of documents through their content (taxonomy creation) and seeking relationships between the content of multiple texts and then setting about linking this information together to form a testable hypothesis about new information. This paper defines and explores an Expertise Seeking Model and the contribution of Text Mining to establish an expertise taxonomy so as to help to compute several weights measuring the experts' affinities and their propensity for collaboration. Part of this work will be by: surveying WWW browsing patterns (Cohen, Maglio, and Barrett 1998) [2], analysing e-mail traffic (Foner 1997) [3] and any other means where text data mining can be put to good use.