# OTTAWA: Optimal TransporT Adaptive Word Aligner for Hallucination and Omission Translation Errors Detection

arXiv: https://arxiv.org/abs/2406.01919 · HTML: https://mlsanity.com/paper/2406.01919/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 21 of its 50 references that are in the map.
- First posted on arXiv 4 Jun 2024.

## Builds on (most cited first, 21 of 21)

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [PyTorch: An Imperative Style, High-Performance Deep Learning Library](https://mlsanity.com/paper/1912.01703/) (2019)
- [Unsupervised Cross-lingual Representation Learning at Scale](https://mlsanity.com/paper/1911.02116/) (2019)
- [Sinkhorn Distances: Lightspeed Computation of Optimal Transportation Distances](https://mlsanity.com/paper/1306.0895/) (2013)
- [COMET: A Neural Framework for MT Evaluation](https://mlsanity.com/paper/2009.09025/) (2020)
- [Word Alignment by Fine-tuning Embeddings on Parallel Corpora](https://mlsanity.com/paper/2101.08231/) (2021)
- [The Curious Case of Hallucinations in Neural Machine Translation](https://mlsanity.com/paper/2104.06683/) (2021)
- [SimAlign: High Quality Word Alignments without Parallel Training Data using Static and Contextualized Embeddings](https://mlsanity.com/paper/2004.08728/) (2020)
- [Hallucinations in Large Multilingual Translation Models](https://mlsanity.com/paper/2303.16104/) (2023)
- [SONAR: Sentence-Level Multimodal and Language-Agnostic Representations](https://mlsanity.com/paper/2308.11466/) (2023)
- [Detecting and Mitigating Hallucinations in Machine Translation: Model Internal Workings Alone Do Well, Sentence Similarity Even Better](https://mlsanity.com/paper/2212.08597/) (2022)
- [Understanding the Properties of Minimum Bayes Risk Decoding in Neural Machine Translation](https://mlsanity.com/paper/2105.08504/) (2021)
- [Improving Pretrained Cross-Lingual Language Models via Self-Labeled Word Alignment](https://mlsanity.com/paper/2106.06381/) (2021)
- [Multi-Vector Models with Textual Guidance for Fine-Grained Scientific Document Similarity](https://mlsanity.com/paper/2111.08366/) (2021)
- [HalOmi: A Manually Annotated Benchmark for Multilingual Hallucination and Omission Detection in Machine Translation](https://mlsanity.com/paper/2305.11746/) (2023)
- [What does Attention in Neural Machine Translation Pay Attention to?](https://mlsanity.com/paper/1710.03348/) (2017)
- [End-to-End Neural Word Alignment Outperforms GIZA++](https://mlsanity.com/paper/2004.14675/) (2020)
- [A Supervised Word Alignment Method based on Cross-Language Span Prediction using Multilingual BERT](https://mlsanity.com/paper/2004.14516/) (2020)
- [Using Optimal Transport as Alignment Objective for fine-tuning Multilingual Contextualized Embeddings](https://mlsanity.com/paper/2110.02887/) (2021)
- [Neural semi-Markov CRF for Monolingual Word Alignment](https://mlsanity.com/paper/2106.02569/) (2021)
- [Toward Interpretable Semantic Textual Similarity via Optimal Transport-based Contrastive Sentence Learning](https://mlsanity.com/paper/2202.13196/) (2022)

## Built on by (most cited first, 1 of 1)

- [Scaling Unsupervised Word Alignment to Documents via Structural Constraints](https://mlsanity.com/paper/2608.21023/) (2026)

## Abstract (arXiv)

Recently, there has been considerable attention on detecting hallucinations and omissions in Machine Translation (MT) systems. The two dominant approaches to tackle this task involve analyzing the MT system's internal states or relying on the output of external tools, such as sentence similarity or MT quality estimators. In this work, we introduce OTTAWA, a novel Optimal Transport (OT)-based word aligner specifically designed to enhance the detection of hallucinations and omissions in MT systems. Our approach explicitly models the missing alignments by introducing a "null" vector, for which we propose a novel one-side constrained OT setting to allow an adaptive null alignment. Our approach yields competitive results compared to state-of-the-art methods across 18 language pairs on the HalOmi benchmark. In addition, it shows promising features, such as the ability to distinguish between both error types and perform word-level detection without accessing the MT system's internal states.