Year: 2026
Type: conference-paper
Abstract: This paper presents a semi-supervised approach to explore the lexical semantic change of English prepositions using the BERT language model.We adopt a method that leverages BERT to obtain representations of prepositions, train a model on these representations, and use this model to measure the lexical semantic changes of the prepositions.We show that BERT embeddings can be used for preposition-sense disambiguation.Then, by applying the trained model to sentences extracted from the Corpus of Historical American English (COHA), we detect a variety of lexical semantic changes over time for English prepositions.
Author: Liudmila Radchankava
Institution: Johannes Gutenberg University Mainz
Topic: Topic Modeling
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: gold
OpenAlex ID: W7140121667
FWCI:
Cites:
Cited by:
Related to: