Generating hierarchical fuzzy concepts from large databases

A concept hierarchy is a kind of concise and general form of knowledge representations. Concept description is vague for human knowledge generally. Crisp description for a concept usually cannot represent human knowledge completely and practically. In this paper, we would study fuzzy characteristics of concept description and propose an agglomerative clustering scheme based on fuzzy theory to generate hierarchical fuzzy concepts from a large database automatically. The proposed method first transforms quantitative data into linguistic terms using fuzzy membership functions. The fuzzy entropy is then designed for evaluating the significant order of attributes and a clustering algorithm is developed to find meaningful fuzzy concept hierarchies effectively.