# Low-Resource Corpus Filtering using Multilingual Sentence Embeddings

arXiv: https://arxiv.org/abs/1906.08885 · HTML: https://mlsanity.com/paper/1906.08885/

## Key facts

- Cited by 25 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 20 Jun 2019.

## Built on by (most cited first, 25 of 25)

- [LLMs Are Few-Shot In-Context Low-Resource Language Learners](https://mlsanity.com/paper/2403.16512/) (2024)
- [A Shocking Amount of the Web is Machine Translated: Insights from Multi-Way Parallelism](https://mlsanity.com/paper/2401.05749/) (2024)
- [There's no Data Like Better Data: Using QE Metrics for MT Data Filtering](https://mlsanity.com/paper/2311.05350/) (2023)
- [Data Selection Curriculum for Neural Machine Translation](https://mlsanity.com/paper/2203.13867/) (2022)
- [Leveraging Multi-lingual Positive Instances in Contrastive Learning to Improve Sentence Embedding](https://mlsanity.com/paper/2309.08929/) (2023)
- [Adapters for Altering LLM Vocabularies: What Languages Benefit the Most?](https://mlsanity.com/paper/2410.09644/) (2024)
- [Self-Supervised Knowledge Assimilation for Expert-Layman Text Style Transfer](https://mlsanity.com/paper/2110.02950/) (2021)
- [Why Not Transform Chat Large Language Models to Non-English?](https://mlsanity.com/paper/2405.13923/) (2024)
- [Integrating Unsupervised Data Generation into Self-Supervised Neural Machine Translation for Low-Resource Languages](https://mlsanity.com/paper/2107.08772/) (2021)
- [How to Learn in a Noisy World? Self-Correcting the Real-World Data Noise in Machine Translation](https://mlsanity.com/paper/2407.02208/) (2024)
- [LAWDR: Language-Agnostic Weighted Document Representations from Pre-trained Models](https://mlsanity.com/paper/2106.03379/) (2021)
- [Exploiting Parallel Corpora to Improve Multilingual Embedding based Document and Sentence Alignment](https://mlsanity.com/paper/2106.06766/) (2021)
- [BitextEdit: Automatic Bitext Editing for Improved Low-Resource Machine Translation](https://mlsanity.com/paper/2111.06787/) (2021)
- [Improving the quality of Web-mined Parallel Corpora of Low-Resource Languages using Debiasing Heuristics](https://mlsanity.com/paper/2502.19074/) (2025)
- [Learning Feature Weights using Reward Modeling for Denoising Parallel Corpora](https://mlsanity.com/paper/2103.06968/) (2021)
- [A Commonsense-Infused Language-Agnostic Learning Framework for Enhancing Prediction of Political Polarity in Multilingual News Headlines](https://mlsanity.com/paper/2212.00298/) (2022)
- [Noisy Parallel Data Alignment](https://mlsanity.com/paper/2301.09685/) (2023)
- [Separating the Wheat from the Chaff with BREAD: An open-source benchmark and metrics to detect redundancy in text](https://mlsanity.com/paper/2311.06440/) (2023)
- [End-to-End Speech Translation for Low-Resource Languages Using Weakly Labeled Data](https://mlsanity.com/paper/2506.16251/) (2025)
- [Improve Sentence Alignment by Divide-and-conquer](https://mlsanity.com/paper/2201.06907/) (2022)
- [Translation Aligned Sentence Embeddings for Turkish Language](https://mlsanity.com/paper/2311.09748/) (2023)
- [A comparison of data filtering techniques for English-Polish LLM-based machine translation in the biomedical domain](https://mlsanity.com/paper/2501.16533/) (2025)
- [UniBERT: Adversarial Training for Language-Universal Representations](https://mlsanity.com/paper/2503.12608/) (2025)
- [From Scarcity to Efficiency: Investigating the Effects of Data Augmentation on African Machine Translation](https://mlsanity.com/paper/2509.07471/) (2025)
- [Leveraging Audio-LLMs to Filter Speech-to-Speech Training Data](https://mlsanity.com/paper/2606.13507/) (2026)

## Abstract (arXiv)

In this paper, we describe our submission to the WMT19 low-resource parallel corpus filtering shared task. Our main approach is based on the LASER toolkit (Language-Agnostic SEntence Representations), which uses an encoder-decoder architecture trained on a parallel corpus to obtain multilingual sentence representations. We then use the representations directly to score and filter the noisy parallel sentences without additionally training a scoring function. We contrast our approach to other promising methods and show that LASER yields strong results. Finally, we produce an ensemble of different scoring methods and obtain additional gains. Our submission achieved the best overall performance for both the Nepali-English and Sinhala-English 1M tasks by a margin of 1.3 and 1.4 BLEU respectively, as compared to the second best systems. Moreover, our experiments show that this technique is promising for low and even no-resource scenarios.