From Semantic Metrics to Semantic Fields
Maksym O. Vakulenko · 2021 IEEE 16th International Conference on Computer Sciences and Information Technologies (CSIT) · 2021
A new approach to build semantic fields based on the metric method to calculate semantic distances between lexical units is put forward. The general case of different weights of semes contributing to the word meaning is considered. It is assumed that the word semes are fed from their definitions or explanations available in a sufficiently large and comprehensive dictionary. We repeatedly apply the formula for semantic representation of a given word treating its semes as new words having their own semes, thus forming a lexicographic semantic tree. Due to hypercycles rising in real dictionaries, the number of iterations is finite which allows one to exactly calculate the weights of semes and, henceforth, to build semantic fields. It is demonstrated that the comprehensive componential analysis of the word should be based on the semantic fields. The described approach may be useful in a variety of Natural Language Processing tasks based on supervised learning that require semantic information: computer lexicography, semantic analysis, information search and retrieval, Machine Translation, etc.