The design and evaluation of EKE,a semi-automated email knowledge extraction tool

This paper presents an approach to locating experts within organisations through the use of the indispensable communication medium and source of information, email. The approach was realised through the email expert locator architecture developed by the authors, which uses email content in the modelling of individuals’ expertise profiles. The approach has been applied to a real-world application, EKE, and evaluated using focus group sessions and system trials. In this work, the authors report the findings obtained from the focus groups sessions. The aim of the sessions was to obtain information about the participants’ perceptions, opinions, underlying attitudes, and recommendations with regard to the notion of exploiting email content for expertise profiling. The paper provides a review of the various approaches to expertise location that have been developed and highlights the end-users’ perspectives on the usability and functionality of EKE and the socio-ethical challenges raised by its adoption from an industrial perspective.