Compressed models for co-reference resolution: enhancing efficiency with debiased word embeddings
Abstract This work presents a comprehensive approach to reduce bias in word embedding vectors and evaluate the impact on various Natural Language Processing (NLP) tasks. Two GloVe variations (840B and 50) are debiased by identifying the gender direction in the word embedding space and then removing...
| Autori principali: | , , , , |
|---|---|
| Natura: | Articolo |
| Lingua: | English |
| Pubblicazione: |
Nature Portfolio
2023-10-01
|
| Serie: | Scientific Reports |
| Accesso online: | https://doi.org/10.1038/s41598-023-45677-0 |