Inferring Adjective Hypernyms with Language Models to Increase the Connectivity of Open English Wordnet

TL;DR

This paper enhances the Open English Wordnet by inferring missing adjective hypernyms using fine-tuned language models, thereby increasing its connectivity and utility in linguistic data applications.

Contribution

It introduces a new resource for adjective hypernymy and adapts the TaxoLLaMa methodology for this specific task, addressing gaps in existing lexical databases.

Findings

Successful fine-tuning of language models for adjective hypernymy prediction

Improved connectivity in Open English Wordnet

Theoretical insights into adjective hypernymy relations

Abstract

Open English Wordnet is a key resource published in OntoLex-lemon as part of the linguistic linked open data cloud. There are, however, many links missing in the resource, and in this paper, we look at how we can establish hypernymy between adjectives. We present a theoretical discussion of the hypernymy relation and how it differs for adjectives in contrast to nouns and verbs. We develop a new resource for adjective hypernymy and fine-tune large language models to predict adjective hypernymy, showing that the methodology of TaxoLLaMa can be adapted to this task.