Comparing the structure of entity nouns between Vietnamese WordNet (VietNet) and English WordNet

Abstract

WordNet is a well-known conceptual system (a type of a synonym dictionary, or a thesaurus) designed for the English language, which has been widely recognized and applied in research works related to computational semantics, lexical semantics, contrastive semantics, etc. VietNet is a similar conceptual system for the Vietnamese language, developed by the Computational Linguistics Center (CLC). To evaluate the structure of VietNet, we need to compare the structure of VietNet with WordNet by aligning the measures in VietNet and WordNet. In this study, the Resnik semantic similarity measure was used to compute semantic distances for 1400 pairs of Vietnamese entity nouns and 1400 pairs of English entity nouns. The comparison results showed that the semantic distance values between Vietnamese and English entity nouns are relatively close, with a small difference of 0.26. This indicates that the structure of VietNet is reliable for entity nouns, as it exhibits a high similarity with the structure of WordNet. The findings of this research are significant and can be applied in various linguistic domains, such as English-Vietnamese comparison in terms of lexical semantics and cross-lingual text similarity. Additionally, the research results highlight the potential of VietNet and the Resnik measure in comparing the semantic distance values of nouns across languages.