A Meta-Level Elicitation and Analysis of Knowledge Requirements

摘要

Knowledge Requirements Elicitation and Analysis (KREA) are two vital tasks in various knowledge-based models and systems. Pure interview-based methods may not be adequate because it is both time-consuming and impractical that KBS requests users to self-report all relevant knowledge requirements. In this paper, we present a meta-level method for eliciting and analysing knowledge requirements. The KREA method consists of two key components: a set of strategies which can extend and formalize user-reported potential knowledge requirements so as to cover implicit or tacit knowledge requirements, and an algorithm for eliminating conflicts among knowledge requirements.