Attribute Reduction in Double Hierarchy Fuzzy Linguistic Formal Context

Attribute reduction is an effective data preprocessing method for knowledge discovery using formal concept analysis. There exists an amount of linguistic evaluation information when experts evaluate different options in uncertain environments. To reduce the complexity in the process of knowledge discovery with linguistic evaluation information, the method of attribute reduction in double hierarchy fuzzy linguistic formal context is discussed. The double hierarchy linguistic term set is first introduced into formal concept analysis to express more detail and complex linguistic evaluation information. The double hierarchy fuzzy linguistic concept lattice is constructed. To simplify knowledge representation, the granular consistent set and granular reduct of double hierarchy fuzzy linguistic formal context are then discussed. A method of attribute reduction based on granular discernibility matrix is developed. The practicality and effectiveness of the proposed method are illustrated through a real-world application example.