A Batch Constructing Method of Weighted Concept Lattice Based on Deviance Analysis

Concept lattice is an effective formal tool for data expression and knowledge acquisition. Weighted concept lattice (WCL) is a structure of concept lattice which depicts the importance of intents. A batch constructing method of weighted concept lattice based on deviance analysis is presented in this paper. The importance deviance value among the multi-attribute intent is computed by the standard deviation. According to the importance threshold and importance deviance threshold of the intent which are given by the user, strongly frequent weighted nodes and their edges are generated from bottom to top, so that the time and storage complexity of constructing weighted concept lattice is reduced, and the practicability and pertinence of concept lattice are improved. The experiment results validate the correctness and validity of the algorithm with the celestial spectrum data as the formal context.