Towards a methodology to elicit tacit domain knowledge from users

Abstract

This paper seeks to address a problem ubiquitous in many software development environments today, namely, building software from requirements that are incomplete and not fully understood, thereby creating products that are either faulty or ultimately not being used at all. This gap that exists between software engineers and clients is highlighted in this paper and suggestions on how to overcome the identified gap are presented. The proposed methodology is to introduce developers into the client's environment, which can be more time consuming and more resource intensive than traditional knowledge elicitation methods, but has the potential to satisfy more of a user's needs in the long run. It also does not seek to replace any of the existing elicitation methods; rather it is complementary to knowledge elicitation techniques currently used by software engineers as well as to enhance current understanding of such processes.