Cross-lingual retrieval aims to retrieve relevant text across languages. Current methods typically achieve cross-lingual retrieval by learning language-agnostic text representations in word or sentence level. However, how to learn phrase representations for cross-lingual phrase retrieval is still an open problem. In this paper, we propose XPR, a cross-lingual phrase retriever that extracts phrase representations from unlabeled example sentences. Moreover, we create a large-scale cross-lingual phrase retrieval dataset, which contains 65K bilingual phrase pairs and 4.2M example sentences in 8 English-centric language pairs. Experimental results show that XPR outperforms state-of-the-art baselines which utilize word-level or sentence-level representations. XPR also shows impressive zero-shot transferability that enables the model to perform retrieval in an unseen language pair during training. Our dataset, code, and trained models are publicly available at www.github.com/cwszz/XPR/.
@article{arxiv.2204.08887,
title = {Cross-Lingual Phrase Retrieval},
author = {Heqi Zheng and Xiao Zhang and Zewen Chi and Heyan Huang and Tan Yan and Tian Lan and Wei Wei and Xian-Ling Mao},
journal= {arXiv preprint arXiv:2204.08887},
year = {2022}
}