Carolyn Mair (Southampton Solent University)
Martin Shepperd (Brunel University)
Mark Stephens (EDS, UK)
Analogical or case-based reasoning (CBR) is a knowledge management technology based upon problem-solving using episodic memory and retrieval by similarity. It has been used for many problems including support for software project management in areas such as prediction and lessons learned. Results from using CBR range from ‘extremely impressive’ to ‘weaker than benchmark' and it is unclear why this should be. Consequently we cannot predict a priori when CBR will aid problem solving. Hence, the aim of this project is to empirically investigate the cognitive processes of professionals using CBR tools for problem-solving (presently almost completely unexplored) in order to develop more effective CBR support for software engineering problems. We will work with EDS, UK to generate realistic problems drawn from the area of project effort prediction. From this we will conduct empirical studies with professional participants using our existing CBR shell, archANGEL. We will use multiple techniques including interviews and think-aloud protocols based on a Grounded Theory approach. This project will lead to (i) improved understanding and therefore utilization of analogy-based project management tools (CBR) and (ii) recommendations for more effective future CBR tools. This research is important because CBR is an increasingly used technology, yet not always effectively deployed, because we have little understanding of how professionals solve problems using analogy-based tools.Key words: cognitive, problem solving, CBR, empirical software engineering, software project estimation.