Abstract
Words in different languages rarely cover the exact same semantic space. This work characterizes differences in meaning between words across languages using semantic relations that have been used to relate the meaning of English words. However, because of translation ambiguity, semantic relations are not always preserved by translation. We introduce a cross-lingual relation classifier trained only with English examples and a bilingual dictionary. Our classifier relies on a novel attention-based distillation approach to account for translation ambiguity when transferring knowledge from English to cross-lingual settings. On new English-Chinese and English-Hindi test sets, the resulting models largely outperform baselines that more naively rely on bilingual embeddings or dictionaries for cross-lingual transfer, and approach the performance of fully supervised systems on English tasks.- Anthology ID:
- D19-1532
- Volume:
- Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)
- Month:
- November
- Year:
- 2019
- Address:
- Hong Kong, China
- Venues:
- EMNLP | IJCNLP
- SIG:
- SIGDAT
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 5285–5296
- Language:
- URL:
- https://aclanthology.org/D19-1532
- DOI:
- 10.18653/v1/D19-1532
- Cite (ACL):
- Yogarshi Vyas and Marine Carpuat. 2019. Weakly Supervised Cross-lingual Semantic Relation Classification via Knowledge Distillation. In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP), pages 5285–5296, Hong Kong, China. Association for Computational Linguistics.
- Cite (Informal):
- Weakly Supervised Cross-lingual Semantic Relation Classification via Knowledge Distillation (Vyas & Carpuat, EMNLP-IJCNLP 2019)
- PDF:
- https://preview.aclanthology.org/remove-xml-comments/D19-1532.pdf