Knowledge Acquisition Via Multi-perspective Concept Formation

Joao Jose Furtado VascoColette FaucherEugene Chouraqui

In this paper, we describe an architecture for helping in the construction of concept hierarchies. This architecture is based on machine learning and on cognitive psychology studies in concept formation. Our basic assumption is that concept formation should be considered as the goal-driven, context-dependent process and, therefore, that the hierarchical organization of concepts should be represented in different perspectives. The core of our architecture is a learning system that generates multi-perspective hierarchies. The evaluation of the architecture is realized from a perspective of both comprehesibility and prediction power of the generated knowledge.

Caso o link acima esteja inválido, faça uma busca pelo texto completo na Web: Buscar na Web

Biblioteca Digital Brasileira de Computação - Contato:
     Mantida por: