The development of genetic algorithm for semantic similarity estimation in terms of knowledge management problems

Abstract

This article is devoted to the development of a new approach for semantic similarity estimation. The main problem in knowledge search field is the complexity of identification and usage of key information, which is increasing constantly. To solve this problem we propose to modify previously developed knowledge filter running on the basis of the semantic concepts taxonomy tree as a systematization of complex areas and hierarchical knowledge. The knowledge filter meta-model is supplemented by a semantic similarity estimation block to obtain the most appropriate results in the context of semantics. We analyzed the assigned problem and observed different ways of semantic similarity estimation. To solve the problem we propose the graph model containing components of ontology triplets. The semantic similarity formula is presented in this paper. To increase the efficiency we developed the genetic algorithm for semantic similarity estimation. Experiments carried out on benchmarks show the efficiency of developed approach.