Efficient management of concepts and properties in conceptual spaces using data mining techniques

Abstract

A conceptual space provides a robust infrastructure for many cognitive tasks including reasoning, perception, action, induction and semantics. Concepts and properties are two key elements in conceptual spaces. Efficient management of these primitives is a fundamental task and opens a road to the future applications. In this paper, we propose two mappings that build conceptual relational tables to which we can apply association rules mining approaches. Proposed approaches efficiently identify frequent patterns and intercorrelated domains that lead to effective management of concepts and properties in data-rich conceptual spaces.