An Evaluation of Expertise in a File Search Environment

Abstract

The evaluation of expertise is an important activity that should be undertaken when human-computer systems are designed. This activity is extremely critical in systems which automatically provide assistance to inexperienced users. However, the representation of expertise is of equal importance in automated assistance systems. An experiment was designed to address the issues of representation and evaluation of expertise in a file search environment using expert and novice subjects. A target profile methodology was used for developing a file search assistant based on expert behavior. The model of expertise generated from this methodology proved to be capable of capturing expert search strategies. It also was capable of providing a framework for the development of an interactive file search assistant. The investigation points to further research in diagnostic and remedial variables in assistance, as well as adaptive assistance.