# Massively Multilingual Document Alignment with Cross-lingual Sentence-Mover's Distance

arXiv: https://arxiv.org/abs/2002.00761 · HTML: https://mlsanity.com/paper/2002.00761/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 31 Jan 2020.

## Built on by (most cited first, 7 of 7)

- [Quality Estimation without Human-labeled Data](https://mlsanity.com/paper/2102.04020/) (2021)
- [kNN-Embed: Locally Smoothed Embedding Mixtures For Multi-interest Candidate Retrieval](https://mlsanity.com/paper/2205.06205/) (2022)
- [Exploiting Parallel Corpora to Improve Multilingual Embedding based Document and Sentence Alignment](https://mlsanity.com/paper/2106.06766/) (2021)
- [Domain Adaptation of Machine Translation with Crowdworkers](https://mlsanity.com/paper/2210.15861/) (2022)
- [Metric Learning in Multilingual Sentence Similarity Measurement for Document Alignment](https://mlsanity.com/paper/2108.09495/) (2021)
- [CDA: a Cost Efficient Content-based Multilingual Web Document Aligner](https://mlsanity.com/paper/2102.10246/) (2021)
- [A General-Purpose Multilingual Document Encoder](https://mlsanity.com/paper/2305.07016/) (2023)

## Abstract (arXiv)

Document alignment aims to identify pairs of documents in two distinct languages that are of comparable content or translations of each other. Such aligned data can be used for a variety of NLP tasks from training cross-lingual representations to mining parallel data for machine translation. In this paper we develop an unsupervised scoring function that leverages cross-lingual sentence embeddings to compute the semantic distance between documents in different languages. These semantic distances are then used to guide a document alignment algorithm to properly pair cross-lingual web documents across a variety of low, mid, and high-resource language pairs. Recognizing that our proposed scoring function and other state of the art methods are computationally intractable for long web documents, we utilize a more tractable greedy algorithm that performs comparably. We experimentally demonstrate that our distance metric performs better alignment than current baselines outperforming them by 7% on high-resource language pairs, 15% on mid-resource language pairs, and 22% on low-resource language pairs.