| Title: | Mono- and cross-lingual evaluation of representation language models on less-resourced languages |
|---|
| Authors: | ID Ulčar, Matej (Author)
ID Žagar, Aleš (Author)
ID Armendariz, Carlos S. (Author)
ID Repar, Andraž, Institut "Jožef Stefan" (Author)
ID Pollak, Senja, Institut "Jožef Stefan" (Author)
ID Purver, Matthew, Institut "Jožef Stefan" (Author)
ID Robnik Šikonja, Marko (Author) |
|---|
| Files: |  URL - Source URL, visit https://www.sciencedirect.com/science/article/pii/S0885230825000774?via%3Dihub
 PDF - Presentation file, download (2,28 MB)
MD5: B4BE1F405393A7D919CA1A369BFD46BC
|
|---|
| Language: | English |
|---|
| Typology: | 1.01 - Original Scientific Article |
|---|
| Organization: |  IJS - Jožef Stefan Institute
|
|---|
| Abstract: | The current dominance of large language models in natural language processing is based on their contextual awareness. For text classification, text representation models, such as ELMo, BERT, and BERT derivatives, are typically fine-tuned for a specific problem. Most existing work focuses on English; in contrast, we present a large-scale multilingual empirical comparison of several monolingual and multilingual ELMo and BERT models using 14 classification tasks in nine languages. The results show, that the choice of best model largely depends on the task and language used, especially in a cross-lingual setting. In monolingual settings, monolingual BERT models tend to perform the best among BERT models. Among ELMo models, the ones trained on large corpora dominate. Cross-lingual knowledge transfer is feasible on most tasks already in a zero-shot setting without losing much performance. |
|---|
| Keywords: | monolingual models, multilingual models, ELMo, BERT, corpus, cross-lingual datasets |
|---|
| Publication status: | Published |
|---|
| Publication version: | Version of Record |
|---|
| Submitted for review: | 03.09.2023 |
|---|
| Article acceptance date: | 03.06.2025 |
|---|
| Publication date: | 27.06.2025 |
|---|
| Publisher: | Elsevier |
|---|
| Year of publishing: | 2026 |
|---|
| Number of pages: | 1-29 str. |
|---|
| Numbering: | Vol. 95, [article no.] 101852 |
|---|
| Source: | Nizozemska |
|---|
| PID: | 20.500.12556/DiRROS-22874  |
|---|
| UDC: | 004.8 |
|---|
| ISSN on article: | 1095-8363 |
|---|
| DOI: | 10.1016/j.csl.2025.101852 |
|---|
| COBISS.SI-ID: | 241622275 |
|---|
| Copyright: | © 2025 The Authors. |
|---|
| Note: | Nasl. z nasl. zaslona;
Opis vira z dne 7. 7. 2025;
|
|---|
| Publication date in DiRROS: | 07.07.2025 |
|---|
| Views: | 326 |
|---|
| Downloads: | 214 |
|---|
| Metadata: |    |
|---|
|
:
|
Copy citation |
|---|
| | | | Share: |  |
|---|
Hover the mouse pointer over a document title to show the abstract or click
on the title to get all document metadata. |
