Pre-empting user questions through anticipation: data mining FAQ lists
Dick Ng’ambi · South African Institute of Computer Scientists and Information Technologists · 2002
In this paper we describe the use of data mining techniques on frequently referenced questions (FRQ) to predict the user's 'next' question with the view to pre-empting the question using proactive response. Relationships and patterns hidden in frequently asked questions (FAQ) lists, once discovered, can be used to anticipate user questions and enrich the questioning engagement. A prototype, dynamic Intelligent Handler of Frequently Asked Questions, has been developed to help predict user questions through knowledge discovery from FAQ lists. This paper reports on the motivation for this work and describes the question prediction technique. FAQ lists from the ITCS (Information Technology and Computing Services) help desk of the University of Cape Town are used to illustrate the technique.