United States

This paper presents an evaluation of 16 machine translation systems submitted to the Shared Task of the 9th Conference of Machine Translation (WMT24) for the English-Hindi (en-hi) language pair using our Complex Structures Test (CoST) suite. Aligning with this year’s test suite sub- task theme, “Help us break LLMs”, we curated a comprehensive test suite encompassing diverse datasets across various categories, including autobiography, poetry, legal, conversation, play, narration, technical, and mixed genres. Our evaluation reveals that all the systems struggle significantly with the archaic style of text like legal and technical writings or text with creative twist like conversation and poetry datasets, highlighting their weaknesses in handling complex linguistic structures and stylistic nuances inherent in these text types. Our evaluation identifies the strengths and limitations of the submitted models, pointing to specific areas where further research and development are needed to enhance their performance.

EMNLP 2024

CoST of breaking the LLMs

breaking the llms

mt evaluation

test suite

challenge set

## Welcome to EMNLP 2024! 
We are excited to welcome you to one of the most prominent conferences in the field of Natural Language Processing. This year, EMNLP 2024 is being held in a hybrid format,
offering both virtual and in-person participation in beautiful Miami. Due to a record-breaking number of submissions, we've expanded the total number of accepted papers to accommodate more cutting-edge research from around the globe.
### [Conference Handbook](https://drive.google.com/file/d/1WPROgxjLAC96AJL7Ugy0tEnYm7dkrbHt/view?usp=sharing)

You are required to register for this event. **Please register [here](https://2024.emnlp.org/registration/).** The EMNLP 2024 event page on Underline will be open to public one week prior to the event.

Please register!

EMNLP 2024 will take place in Miami, Florida from Nov 12th to Nov 16th, 2024, at the Hyatt Regency Miami Hote and on Underline for remote participants.

We assess the difficulty of gender resolution in literary-style dialogue settings and the influence of gender stereotypes. Instances of the test suite contain spoken dialogue interleaved with external meta-context about the characters and the manner of speaking. We find that character and manner stereotypes outside of the dialogue significantly impact the gender agreement of referents within the dialogue.

WMT24 Test Suite: Gender Resolution in Speaker-Listener Dialogue Roles

The GenderQueer Test Suite is designed to probe MT systems’ capabilities and limitations in handling gender agreement in morphologically rich languages such as Icelandic as well as non-binary pronouns and LGBTQIA+ terminology.

The GenderQueer Test Suite

This paper summarizes the results of our test suite evaluation on 39 machine translation systems submitted at the Shared Task of the Ninth Conference of Machine Translation (WMT24). It offers a fine-grained linguistic evaluation of machine translation outputs for English– German and English–Russian, resulting from a significant manual linguistic effort. Based on our results, LLMs are inferior to NMT in English–German, both in overall scores and when translating specific linguistic phenomena, such as punctuation, complex future verb tenses, and stripping. LLMs show quite a competitive performance in English-Russian, although top-performing systems might struggle with some cases of named entities and terminology, function words, mediopassive voice, and semantic roles. Additionally, some LLMs generate very verbose or empty outputs, posing challenges to the evaluation process.

Investigating the Linguistic Performance of Large Language Models in Machine Translation

We present MetaMetrics-MT, an innovative metric designed to evaluate machine translation (MT) tasks by aligning closely with human preferences through Bayesian optimization with Gaussian Processes. MetaMetrics-MT enhances existing MT metrics by optimizing their correlation with human judgments. Our experiments on the WMT24 metric shared task dataset demonstrate that MetaMetrics-MT outperforms all existing baselines, setting a new benchmark for state-of-the-art performance in the referencebased setting. Furthermore, it achieves comparable results to leading metrics in the referencefree setting, offering greater efficiency.

MetaMetrics-MT: Tuning Meta-Metrics for Machine Translation via Human Preference Calibration

Machine translation (MT) evaluation metrics like BLEU and chrF++ are widely used reference-based metrics that do not require training and are language-independent. How- ever, these metrics primarily focus on n- gram matching and often overlook semantic depth and contextual understanding. To ad- dress this gap, we introduce chrF-S (Seman- tic chrF++), an enhanced metric that integrates sentence embeddings to evaluate translation quality more comprehensively. By combining traditional character and word n-gram analysis with semantic information derived from em- beddings, chrF-S captures both syntactic ac- curacy and sentence-level semantics. This pa- per presents our contributions to the WMT24 shared metrics task, showcasing our partici- pation and the development of chrF-S. We also demonstrate that, according to prelimi- nary results on the leaderboard, our metric performs on par with other supervised and LLM-based metrics. By merging semantic in- sights with n-gram precision, chrF-S offers a significant enhancement in the assessment of machine-generated translations, advancing the field of MT evaluation.

chrF-S: Semantics Is All You Need

In this paper, we present the MetricX-24 submissions to the WMT24 Metrics Shared Task and provide details on the improvements we made over the previous version of MetricX. Our primary submission is a hybrid reference-based/-free metric, which can score a translation irrespective of whether it is given the source segment, the reference, or both. The metric is trained on previous WMT data in a two-stage fashion, first on the DA ratings only, then on a mixture of MQM and DA ratings. The training set in both stages is augmented with synthetic examples that we created to make the metric more robust to several common failure modes, such as fluent but unrelated translation, or undertranslation. We demonstrate the benefits of the individual modifications via an ablation study, and show a significant performance increase over MetricX-23 on the WMT23 MQM ratings, as well as our new synthetic challenge set.

MetricX-24: The Google Submission to the WMT 2024 Metrics Shared Task

In machine translation quality estimation (QE), translation quality is evaluated automatically without the need for reference translations. This paper describes our contribution to the sentence-level subtask of Task 1 at the Ninth Machine Translation Conference (WMT24), which predicts quality scores for neural MT outputs without reference translations. We fine-tune GPT-4o mini, a large-scale language model (LLM), with limited data for QE. We report results for the direct assessment (DA) method for four language pairs: English-Gujarati (En-Gu), English-Hindi (En-Hi), English-Tamil (En-Ta), and English-Telugu (En-Te). Experiments under zero-shot, few-shot prompting, and fine-tuning settings revealed significantly low performance in the zero-shot, while fine-tuning achieved accuracy comparable to last year’s best scores. Our system demonstrated the effectiveness of this approach in low-resource language QE, securing 1st place in both En-Gu and En-Hi, and 4th place in En-Ta and En-Te. The code used in our experiments is available at the following URL (https://colab.research.google.com/drive/1p8VMnAkRfuVpbvM_revV2ZaN76sSxmiE?usp=sharing)..

TMU-HIT’s Submission for the WMT24 Quality Estimation Shared Task: Is GPT-4 a Good Evaluator for Machine Translation?

As part of the Open Language Data Initiative shared tasks, we have expanded the FLORES+ evaluation set to include Emakhuwa, a lowresource language widely spoken in Mozambique. We translated the dev and devtest sets from Portuguese into Emakhuwa, and we detail the translation process and quality assurance measures used. Our methodology involved various quality checks, including postediting and adequacy assessments. The resulting datasets consist of multiple reference sentences for each source. We present baseline results from training a Neural Machine Translation system and fine-tuning existing multilingual translation models. Our findings suggest that spelling inconsistencies remain a challenge in Emakhuwa. Additionally, the baseline models underperformed on this evaluation set, underscoring the necessity for further research to enhance machine translation quality for Emakhuwa. The data is publicly available at https://huggingface.co/ datasets/LIACC/Emakhuwa-FLORES

Expanding FLORES+ Benchmark for More Low-Resource Settings: Portuguese-Emakhuwa Machine Translation Evaluation

We introduce a FLORES+ dataset as an evaluation benchmark for modern Wu Chinese machine translation models and showcase its compatibility with existing Wu data. Wu Chinese is mutually unintelligible with other Sinitic languages such as Mandarin and Yue (Cantonese), but uses a set of Hanzi (Chinese characters) that profoundly overlaps with others. The population of Wu speakers is the second largest among languages in China, but the language has been suffering from significant drop in usage especially among the younger generations. We identify Wu Chinese as a textually low-resource language and address challenges for its machine translation models. Our contributions include: (1) an open-source, manually translated dataset, (2) full documentations on the process of dataset creation and validation experiments, (3) preliminary tools for Wu Chinese normalization and segmentation, and (4) benefits and limitations of our dataset, as well as implications to other low-resource languages.

Machine Translation Evaluation Benchmark for Wu Chinese: Workflow and Analysis

This paper introduces a translation of the FLORES+ dataset into the endangered Erzya language, with the goal of evaluating machine translation between this language and any of the other 200 languages already included into FLORES+. This translation was carried out as a part of the Open Language Data shared task at WMT24. We also present a benchmark of existing translation models bases on this dataset and a new translation model that achieves the state-of-the-art quality of translation into Erzya from Russian and English.

Downloads

Next from EMNLP 2024

WMT24 Test Suite: Gender Resolution in Speaker-Listener Dialogue Roles

Similar lecture

How Grammatical Features Impact Machine Translation: A New Test Suite for Chinese-English MT Evaluation

.css-70qvj9{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;}Downloads

Next from EMNLP 2024

WMT24 Test Suite: Gender Resolution in Speaker-Listener Dialogue Roles

Similar lecture

How Grammatical Features Impact Machine Translation: A New Test Suite for Chinese-English MT Evaluation

Downloads