Building semantic hierarchies of formal concepts by deep cognitive machine learning
Mehrdad Valipour, Yingxu Wang · 2017
Semantic comprehension is a fundamental challenge to machine learning. A methodology for building quantitative semantic hierarchies of formal concepts is formally described by cognitive machine learning. The algorithm of concept semantic hierarchy learning (CSH_Learning) is developed based on a set of semantic analysis and synthesis rules according to concept algebra. The algorithm and rules are applied in quantitative determination of levels of arbitrary concepts for the semantic hierarchy of a machines' cognitive knowledge bases. Experiments on 600+ formal concepts indicate the effectiveness of the algorithm for a wide range of applications including cognitive language processing, semantic analyses, cognitive machine learning, cognitive computation and computational linguistics.