Multi-layered FCMs applied to context dependent learning

In this paper we propose a knowledge-based system (KBS) that incorporates intelligence using methods of learning relational structures and evidential reasoning. Theories of causality in data and the role that learning paradigms play in generalising data, in particular evidential inductive learning schemes using fuzzy logic are discussed. This work is only the first step to providing an adaptive KBS that attempts to explain and categorise data. Census data are used to illustrate the effectiveness of this structure and the problems that human experts face when classifying data as a means to providing generalised explanations for subsequent human activity. It is envisaged that our approach will prove significant in application areas that involve processes for economic decision rationale and planning