Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2512.18357 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866912779754012672 |
|---|---|
| author | Hribach, El Mokhtar Mechhour, Oussama Elmonstaser, Mohammed Boudouri, Yassine El Kabal, Othmane |
| author_facet | Hribach, El Mokhtar Mechhour, Oussama Elmonstaser, Mohammed Boudouri, Yassine El Kabal, Othmane |
| contents | Acronym Disambiguation (AD) is a fundamental challenge in technical text processing, particularly in specialized sectors where high ambiguity complicates automated analysis. This paper addresses AD within the context of the TextMine'26 competition on French railway documentation. We present DACE (Dynamic Prompting, Retrieval Augmented Generation, Contextual Selection, and Ensemble Aggregation), a framework that enhances Large Language Models through adaptive in-context learning and external domain knowledge injection. By dynamically tailoring prompts to acronym ambiguity and aggregating ensemble predictions, DACE mitigates hallucination and effectively handles low-resource scenarios. Our approach secured the top rank in the competition with an F1 score of 0.9069. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2512_18357 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | DACE For Railway Acronym Disambiguation Hribach, El Mokhtar Mechhour, Oussama Elmonstaser, Mohammed Boudouri, Yassine El Kabal, Othmane Computation and Language Acronym Disambiguation (AD) is a fundamental challenge in technical text processing, particularly in specialized sectors where high ambiguity complicates automated analysis. This paper addresses AD within the context of the TextMine'26 competition on French railway documentation. We present DACE (Dynamic Prompting, Retrieval Augmented Generation, Contextual Selection, and Ensemble Aggregation), a framework that enhances Large Language Models through adaptive in-context learning and external domain knowledge injection. By dynamically tailoring prompts to acronym ambiguity and aggregating ensemble predictions, DACE mitigates hallucination and effectively handles low-resource scenarios. Our approach secured the top rank in the competition with an F1 score of 0.9069. |
| title | DACE For Railway Acronym Disambiguation |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2512.18357 |