Improving decision support systems with a High Stake Community Contributed Knowledge Base

Abstract

In this paper, we present an approach for high stake decision making based on the processing of uncontrolled knowledge originated from e-communities. Knowledge bases fed by such communities can provide a very rich source for a diagnosis or decision-support systems. However, they inherently bear inherent problems due in particular to knowledge heterogeneity, vagueness, completeness, uncertainty, and origins. In the High Stake Community Contributed Knowledge Base (HSCCKB) approach, the challenge of exploiting heterogeneous data in high stake decisions is addressed. We propose here a dedicated knowledge model as well as tracks for the evaluation of important knowledge aspects, and focus on the decisional architecture and process involved in HSCCKB-based decision support systems.