Work

The Impact of Word Splitting on the Semantic Content of Contextualized Word Representations

Year: 2024

Type: article

Abstract: When deriving contextualized word representations from language models, a decision needs to be made on how to obtain one for out-of-vocabulary (OOV) words that are segmented into subwords. What is the best way to represent these words with a single vector, and are these representations of worse quality than those of in-vocabulary words? We carry out an intrinsic evaluation of embeddings from different models on semantic similarity tasks involving OOV words. Our analysis reveals, among other interesting findings, that the quality of representations of words that are split is often, but not always, worse than that of the embeddings of known words. Their similarity values, however, must be interpreted with caution.

Source: Transactions of the Association for Computational Linguistics

DOI: 10.1162/tacl_a_00647

Authors: Aina Garí Soler, Matthieu Labeau, Chloé Clavel

Institutions: Télécom Paris, Institut Polytechnique de Paris, Laboratoire Traitement et Communication de l’Information, Institut national de recherche en sciences et technologies du numérique

Language: English

Topic: Topic Modeling

Subfield: Artificial Intelligence

Field: Computer Science

Domain: Physical Sciences

Open Access status: diamond

Funders: Agence Nationale de la Recherche, Télécom Paris

Award: viRtual bEhaVioral skIlls TrAining for pubLIc SpEaking