Evaluation of the CLEF query interface
Catalina Hallett, Donia Scott, Richard Power · Open Research Online - ORO (The Open University) · 2006
As part of the clef project, we have developed a method for allowing subject matter experts to pose complex queries to a database. The method makes use of natural language generation, whereby users compose queries through interaction with a dynamic text that is generated "on the fly". We present here some first-step evaluations that we have conducted with our modest pool of available testers, the numbers of which are constrained for practical reasons having to do with data security. The results allow us to draw some useful conclusions about the usability of the method, and its training requirements, for subjects that are representative of our "typical" intended end-users.