Challenge: Parallel corpora are key to developing good machine translation systems, but abundant parallel data is hard to come by for languages with a low number of speakers.
Approach: They propose an unsupervised alignment method that can handle rich morphology by removing incorrect translations and segments containing extraneous data.
Outcome: The proposed method maximizes the number of correctly translated segments in a corpus and minimises noise by removing incorrect translations and segments containing extraneous data.

Similar Papers

Not Low-Resource Anymore: Aligner Ensembling, Batch Filtering, and New Datasets for Bengali-English Machine Translation (2020.emnlp-main)

Copied to clipboard

Challenge: despite being the seventh most widely spoken language, Bengali has received little attention in machine translation due to being low in resources.
Approach: They propose a customized sentence segmenter for Bengali and two new methods for parallel corpus creation on low-resource setups.
Outcome: The proposed method improves Bengali-English parallel corpus by 9 BLEU over previous approaches . the results will pave the way for future research on Bengali and other low-resource languages .
Word Alignment by Fine-tuning Embeddings on Parallel Corpora (2021.eacl-main)

Copied to clipboard

Challenge: Existing work on word alignment has focused on unsupervised learning on parallel text.
Approach: They propose to combine pre-trained contextualized word embeddings with multilingually trained language models to achieve competitive results on word alignment tasks.
Outcome: The proposed model outperforms state-of-the-art models on five language pairs and can train multilingual word aligners that can obtain robust performance on different language pairs.
A Recipe of Parallel Corpora Exploitation for Multilingual Large Language Models (2025.findings-naacl)

Copied to clipboard

Challenge: Recent studies have highlighted the potential of exploiting parallel corpora to enhance multilingual large language models.
Approach: They investigate the impact of parallel corpora quality and quantity, training objectives, and model size on performance of multilingual large language models enhanced with parallel corporeal.
Outcome: The proposed approach improves performance in bilingual and general-purpose tasks.
Building Comparable Corpora for Assessing Multi-Word Term Alignment (2022.lrec-1)

Copied to clipboard

Challenge: Existing methods to extract bilingual terminologies from corpora are limited . MWTs pose serious challenges for alignment and machine translation systems .
Approach: They propose an approach to build comparable corpora and bilingual term dictionaries that evaluate bilingual term alignment in comparable corpus.
Outcome: The proposed method is validated on an existing dataset and manually annotated data.
SimAlign: High Quality Word Alignments Without Parallel Training Data Using Static and Contextualized Embeddings (2020.findings-emnlp)

Copied to clipboard

Challenge: Word alignments are useful for statistical and neural machine translation (NMT) and cross-lingual annotation projection.
Approach: They propose to leverage multilingual word embeddings for word alignment.
Outcome: The proposed methods perform better for four languages and comparable for two languages than traditional statistical aligners even with abundant parallel data.
Noisy Parallel Data Alignment (2023.findings-eacl)

Copied to clipboard

Challenge: Optical character recognition (OCR) is used to convert endangered language documents into machine-readable data, but its noisy outputs are a challenge for many under-resourced languages.
Approach: They propose to use optical character recognition (OCR) to convert endangered language documents into machine-readable data by using noisy alignment models.
Outcome: The proposed model reduces alignment error rate on a state-of-the-art neural-based alignment model up to 59.6%.
AlignFix: A Tool for Parallel Corpora Augmentation and Refinement (2026.eacl-demo)

Copied to clipboard

Challenge: High-quality datasets are crucial for training effective state of the art machine translation systems, but they can be noisy and degrade performance.
Approach: They propose an open-source tool for augmenting data, identifying and correcting errors in parallel corpora.
Outcome: The tool extracts consistent phrase pairs, enabling targeted replacements that can improve the dataset quality.
From Unaligned to Aligned: Scaling Multilingual LLMs with Multi-Way Parallel Corpora (2025.emnlp-main)

Copied to clipboard

Challenge: Experiments show that models trained on multi-way parallel data outperform those trained on unaligned data.
Approach: They propose a large-scale, high-quality multi-way parallel corpus based on TED Talks that spans 113 languages with up to 50 languages aligned in parallel.
Outcome: The proposed model outperforms models trained on unaligned multilingual data on six multilingual benchmarks.
Meeting the Needs of Low-Resource Languages: The Value of Automatic Alignments via Pretrained Models (2023.eacl-main)

Copied to clipboard

Challenge: Large multilingual models have inspired a new class of word alignment methods, which work well for pretraining languages.
Approach: They propose to use transformer-based word alignment methods to extract alignments from massive pretrained models.
Outcome: The proposed methods outperform traditional methods for languages unseen to pretraining models, and are competitive with each other.
OpenSubtitles2018: Statistical Rescoring of Sentence Alignments in Large, Noisy Parallel Corpora (L18-1)

Copied to clipboard

Challenge: Movie and TV subtitles are a valuable resource for the compilation of parallel corpora . however, the quality of the resulting sentence alignments is often lower than for other parallel corpoora.
Approach: They propose to use movie and TV subtitles to extract parallel corpora from 3.7 million subtitles spread over 60 languages to obtain explicit quality scores for each sentence alignment.
Outcome: The proposed model predicts translation probabilities with a root mean square error of 0.07 . the results show that the model can prune out low-quality alignments .

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations