Papers by Raj Dabre

Robustness of Multi-Source MT to Transcription Errors (2023.findings-acl)

Challenge:	IndicNLG is a non-English language that is hampered by the scarcity of datasets.
Approach:	They propose to create a dataset for natural language generation for 11 Indic languages . they use a set of pre-trained models to train multilingual models .
Outcome:	The proposed datasets show that pre-trained models perform well in multilingual and monolingual tasks.

Copied to clipboard

Dominik Macháček, Peter Polák, Ondřej Bojar, Raj Dabre

KreolMorisienMT: A Dataset for Mauritian Creole Machine Translation (2022.findings-aacl)

Challenge:	In multilingual settings, the same content may be available in various languages via simultaneous interpreting, dubbing or subtitling.
Approach:	They hypothesize that leveraging multiple sources will improve translation quality if the sources complement one another in terms of correct information they contain.
Outcome:	The proposed method is robust to speech recognition errors on a 10-hour ESIC corpus.

Copied to clipboard

Raj Dabre, Aneerav Sukhoo

How Good is Zero-Shot MT Evaluation for Low Resource Indian Languages? (2024.acl-short)

Challenge:	Mauritian Creole is a French-based creole and a lingua franca of the Republic of Mauritius.
Approach:	They describe a dataset for benchmarking machine translation quality of Mauritian Creole.
Outcome:	The proposed dataset compares KreolMorisienMT with existing models and human evaluation reveals the systems’ high translation quality.

Copied to clipboard

Anushka Singh, Ananya Sai, Raj Dabre, Ratish Puduppully, Anoop Kunchukuttan, Mitesh Khapra

Low-Resource NLP

Challenge:	a recent study focused on machine translation evaluation for low-resource languages . linguistic aspects that vary across languages are factors that will exacerbate the problem in low-source languages due to the reliance on extensive data resources.
Approach:	They propose to use multi-dimensional quality metrics and DA annotations to meta-evaluate MT evaluation metrics for low-resource languages.
Outcome:	The proposed evaluation metrics are based on human scores on the candidate translations of assamese, maithili, and Punjabi.

Multilingual Neural Machine Translation (2020.coling-tutorials)

Copied to clipboard

Raj Dabre, Chenhui Chu, Anoop Kunchukuttan

The Reasoning Lingua Franca: A Double-Edged Sword for Multilingual AI (2026.eacl-short)

Challenge:	In this tutorial, we will cover the latest advances in NMT to enhance low-resource translation.
Approach:	They will cover the latest advances in NMT approaches that leverage multilingualism . they will focus on topics such as language divergence, transfer learning and pivoting .
Outcome:	This tutorial will cover the latest advances in NMT to enhance low-resource translation models.

Copied to clipboard

Alan Saji, Raj Dabre, Anoop Kunchukuttan, Ratish Puduppully

Scripts Through Time: A Survey of the Evolving Role of Transliteration in NLP (2026.findings-acl)

Challenge:	Large Reasoning Models (LRMs) are highly effective on mathematical, scientific, and other question-answering tasks.
Approach:	They compare an LRM's reasoning in English to that of a multilingual question . they find that English reasoning traces exhibit a substantially higher presence of cognitive behaviors .
Outcome:	The LRMs generate reasoning sequences in English, but the language of the question is not.

Copied to clipboard

Thanmay Jayakumar, Deepon Halder, Raj Dabre

Are Language Models Agnostic to Linguistically Grounded Perturbations? A Case Study of Indic Languages (2025.findings-naacl)

Challenge:	Cross-lingual transfer is often hindered by the "script barrier" where differences in writing systems inhibit transfer learning . transliteration is a powerful technique to bridge this gap by increasing lexical overlap . authors present a taxonomy of key motivations to utilize transliterations in language models .
Approach:	They propose a taxonomy of key motivations to utilize transliterations in NLP . they analyze the evolution and effectiveness of these methods and discuss trade-offs .
Outcome:	The proposed transliteration technique is effective in cross-lingual NLP, the authors argue . the proposed translliteration method is a powerful tool to overcome the "script barrier"

Copied to clipboard

Poulami Ghosh, Raj Dabre, Pushpak Bhattacharyya

Robustness in NLP

Challenge:	Existing studies do not focus on linguistically grounded attacks, but pre-trained models are susceptible to these perturbations.
Approach:	They propose to examine whether pre-trained language models are agnostic to linguistically grounded attacks . they find that PLMs are less susceptible to linguistic perturbations than non-linguistic ones .
Outcome:	The proposed model is agnostic to linguistically grounded attacks, but is less susceptible to linguist attacks than non-linguistic models.

PrahokBART: A Pre-trained Sequence-to-Sequence Model for Khmer Natural Language Generation (2025.coling-main)

Copied to clipboard

Hour Kaing, Raj Dabre, Haiyue Song, Van-Hien Tran, Hideki Tanaka, Masao Utiyama

Data and Model Centric Approaches for Expansion of Large Language Models to New languages (2025.emnlp-tutorials)

Challenge:	Pre-trained sequence-to-sequence models are typically pretrained on extensive raw text corpora and fine-tuned on task-specific data.
Approach:	They introduce a pre-trained sequence-to-sequence model trained from scratch for Khmer using carefully curated Khmer and English corpora.
Outcome:	The proposed model outperforms existing models on three generative tasks and is data-efficient and effective in enhancing performance across various natural language generation tasks.

Copied to clipboard

Anoop Kunchukuttan, Raj Dabre, Rudra Murthy, Mohammed Safi Ur Rahman Khan, Thanmay Jayakumar

A Morphology-Based Investigation of Positional Encodings (2024.emnlp-main)

Challenge:	Existing LLMs mainly support English alongside a handful of high resource languages . this leaves a major gap for most low-resource languages despite increasing pace of research .
Approach:	This tutorial examines approaches to expand the language coverage of LLMs . they look at tokenizer training, pre-training, instruction tuning, alignment, evaluation, etc.
Outcome:	This tutorial examines approaches to expand the language coverage of LLMs . it provides an efficient and viable path to bring LLM technologies to low-resource languages .

Copied to clipboard

Poulami Ghosh, Shikhar Vashishth, Raj Dabre, Pushpak Bhattacharyya

Syntactic Text Processing

Challenge:	Contemporary deep learning models handle languages with diverse morphology . morphological complexity of languages is closely linked with positional encodings .
Approach:	They propose to use positional encodings to integrate morphological complexity into deep learning models.
Outcome:	The proposed model improves on 22 languages and 5 downstream tasks.

YANMTT: Yet Another Neural Machine Translation Toolkit (2023.acl-demo)

Copied to clipboard

Raj Dabre, Diptesh Kanojia, Chinmay Sawant, Eiichiro Sumita

FeatureBART: Feature Based Sequence-to-Sequence Pre-Training for Low-Resource NMT (2022.coling-1)

Challenge:	Neural machine translation (NMT) is an end-to-end approach that provides stateof-the-art results for a variety of language pairs.
Approach:	They propose to build an open-source neural machine translation toolkit on top of HuggingFace's Transformers library and use it for pre-training and fine-tuning sequence-to-sequence models.
Outcome:	The proposed toolkit is built on top of the HuggingFace Transformers library and provides advanced features such as document/multi-source NMT, simultaneous NMT and mixtures-of-experts.

Copied to clipboard

Abhisek Chakrabarty, Raj Dabre, Chenchen Ding, Hideki Tanaka, Masao Utiyama, Eiichiro Sumita

An Empirical Study of In-context Learning in LLMs for Machine Translation (2024.findings-acl)

Challenge:	FeatureBART is a linguistically motivated sequence-to-sequence monolingual pre-training strategy . syntactic features such as lemma, part-of-speech and dependency labels are incorporated into the pre-trained model .
Approach:	They propose a linguistically motivated sequence-to-sequence monolingual pre-training strategy that incorporates syntactic features into the framework.
Outcome:	The proposed model improves translation quality in bilingual and multilingual settings over models that do not use features.

Copied to clipboard

Pranjal Chitale, Jay Gala, Raj Dabre

A Multilingual Multiway Evaluation Data Set for Structured Document Translation of Asian Languages (2022.findings-aacl)

Challenge:	Recent studies focus on optimizing translation quality, with limited attention to understanding specific aspects of ICL that influence the said quality.
Approach:	They conduct the first of its kind, exhaustive study of in-context learning for machine translation (MT) they establish that ICL is primarily example-driven and not instruction-driven .
Outcome:	The proposed model is based on examples and not instruction-driven learning.

Copied to clipboard

Bianka Buschbeck, Raj Dabre, Miriam Exel, Matthias Huck, Patrick Huy, Raphael Rubino, Hideki Tanaka

JASS: Japanese-specific Sequence to Sequence Pre-training for Neural Machine Translation (2020.lrec-1)

Challenge:	a lack of evaluation data sets for structured content limits progress in machine translation . a common use case of machine translation is the translation of structured or formatted documents .
Approach:	They propose a multilingual multiway evaluation data set for machine translation of structured documents of Asian languages Japanese, Korean and Chinese.
Outcome:	The proposed data set is well suited for multilingual evaluation and contains richer annotation tag sets than existing data sets.

Copied to clipboard

Zhuoyuan Mao, Fabien Cromieres, Raj Dabre, Haiyue Song, Sadao Kurohashi

Pretraining Language Models Using Translationese (2024.emnlp-main)

Challenge:	Neural machine translation (NMT) requires large parallel corpora for training robust and high quality models.
Approach:	They propose a Japanese-specific sequence to sequence pre-training alternative to MASS for NMT . they use Japanese as the source or target language to train their models .
Outcome:	The proposed approach can give competitive results over MASS and BRSS, and significantly surpass the individual methods.

Copied to clipboard

Meet Doshi, Raj Dabre, Pushpak Bhattacharyya

Pre-training via Leveraging Assisting Languages for Neural Machine Translation (2020.acl-srw)

Challenge:	a recent study shows that large language models perform well in low-resource languages . a vast majority of languages don't have comparable data as compared to English .
Approach:	They propose to use Translationese as synthetic data for pre-training language models for low-resource languages.
Outcome:	The proposed method reduces performance of LMs trained on clean data in Indian languages . the proposed model performs better in English than in other languages, but is not comparable to English.

Copied to clipboard

Haiyue Song, Raj Dabre, Zhuoyuan Mao, Fei Cheng, Sadao Kurohashi, Eiichiro Sumita

Coursera Corpus Mining and Multistage Fine-Tuning for Improving Lectures Translation (2020.lrec-1)

Challenge:	Sequence-to-sequence (S2S) pre-training with large monolingual data is not always available for the languages of interest (LOI).
Approach:	They propose to use monolingual corpora of other languages to complement the scarce monolingual LOI by script mapping (Chinese to Japanese) . Using only Chinese and French monolinguals, they improve Japanese-English translation quality by up to 8.5 BLEU in low-resource scenarios.
Outcome:	The proposed approach improves Japanese-English translation quality by up to 8.5 BLEU in low-resource scenarios.

Copied to clipboard

Haiyue Song, Raj Dabre, Atsushi Fujita, Sadao Kurohashi

NICT’s participation to WAT 2019: Multilingualism and Multi-step Fine-Tuning for Low Resource NMT (D19-52)

Challenge:	Lectures translation is a case of spoken language translation and there is nil available corpus for this purpose.
Approach:	They propose a framework for mining a parallel corpus from publicly available lectures at Coursera . they use machine translation and cosine similarity over continuous-space sentence representations to determine sentence alignments .
Outcome:	The proposed framework improves translation performance when used with out-of-domain parallel corpora . it also addresses noise in the mined data, and creates high-quality evaluation splits .

Copied to clipboard

Raj Dabre, Eiichiro Sumita

WorldCuisines: A Massive-Scale Benchmark for Multilingual and Multicultural Visual Question Answering on Global Cuisines (2025.naacl-long)

Challenge:	In this paper, we describe our submissions for the following tasks: English–Tamil translation and Russian–Japanese translation.
Approach:	They propose to use multilingual domain adaptation and back-translation to improve translations in Russian–Japanese and English–Tamil.
Outcome:	The proposed techniques perform better in Russian–Japanese and English–Tamil translation tasks.

Copied to clipboard

Genta Indra Winata, Frederikus Hudi, Patrick Amadeus Irawan, David Anugraha, Rifki Afina Putri, Wang Yutong, Adam Nohejl, Ubaidillah Ariq Prathama, Nedjma Ousidhoum, Afifa Amriani, Anar Rzayev, Anirban Das, Ashmari Pramodya, Aulia Adila, Bryan Wilie, Candy Olivia Mawalim, Cheng Ching Lam, Daud Abolade, Emmanuele Chersoni, Enrico Santus, Fariz Ikhwantri, Garry Kuwanto, Hanyang Zhao, Haryo Akbarianto Wibowo, Holy Lovenia, Jan Christian Blaise Cruz, Jan Wira Gotama Putra, Junho Myung, Lucky Susanto, Maria Angelica Riera Machin, Marina Zhukova, Michael Anugraha, Muhammad Farid Adilazuarda, Natasha Christabelle Santosa, Peerat Limkonchotiwat, Raj Dabre, Rio Alexander Audino, Samuel Cahyawijaya, Shi-Xiong Zhang, Stephanie Yulia Salim, Yi Zhou, Yinxuan Gui, David Ifeoluwa Adelani, En-Shiun Annie Lee, Shogo Okada, Ayu Purwarianti, Alham Fikri Aji, Taro Watanabe, Derry Tanti Wijaya, Alice Oh, Chong-Wah Ngo

Visual Data in NLP

Challenge:	Vision Language Models struggle with cultural-specific knowledge, especially in languages other than English and in underrepresented cultural contexts.
Approach:	They propose a visual question answering (VQA) dataset with text-image pairs across 30 languages and dialects and a training dataset.
Outcome:	The proposed model performs better with correct location context, but struggles with adversarial contexts and predicting specific regional cuisines and languages.

CaMMT: Benchmarking Culturally Aware Multimodal Machine Translation (2025.findings-emnlp)

Copied to clipboard

Emilio Villa-Cueva, Sholpan Bolatzhanova, Diana Turmakhan, Kareem Elzeky, Henok Biadglign Ademtew, Alham Fikri Aji, Vladimir Araujo, Israel Abebe Azime, Jinheon Baek, Frederico Belcavello, Fermin Cristobal, Jan Christian Blaise Cruz, Mary Dabre, Raj Dabre, Toqeer Ehsan, Naome A Etori, Fauzan Farooqui, Jiahui Geng, Guido Ivetta, Thanmay Jayakumar, Soyeong Jeong, Zheng Wei Lim, Aishik Mandal, Sofía Martinelli, Mihail Minkov Mihaylov, Daniil Orel, Aniket Pramanick, Sukannya Purkayastha, Israfel Salazar, Haiyue Song, Tiago Timponi Torrent, Debela Desalegn Yadeta, Injy Hamed, Atnafu Lambebo Tonja, Thamar Solorio

Improving Low-Resource NMT through Relevance Based Linguistic Features Incorporation (2020.coling-main)

Challenge:	a human-curated benchmark of over 5,800 triples of images is used to evaluate multimodal translation systems.
Approach:	They introduce a human-curated benchmark of over 5,800 triples of images along with parallel captions in English and regional languages.
Outcome:	The results show that visual context improves translation quality in culturally-specific items .

Copied to clipboard

Abhisek Chakrabarty, Raj Dabre, Chenchen Ding, Masao Utiyama, Eiichiro Sumita

Towards Building Large Scale Datasets and State-of-the-Art Automatic Speech Translation Systems for 14 Indian Languages (2025.acl-long)

Challenge:	Existing studies on incorporating arbitrary syntactic information into neural machine translation (NMT) are lacking.
Approach:	They propose to integrate linguistic knowledge at different levels into neural machine translation framework to improve translation quality for language pairs with extremely limited data.
Outcome:	The proposed methods improve translation quality for all tasks by 3.09 BLEU points . the proposed methods are based on two different approaches .

Copied to clipboard

Ashwin Sankar, Sparsh Jain, Nikhil Narasimhan, Devilal Choudhary, Dhairya Suman, Mohammed Safi Ur Rahman Khan, Anoop Kunchukuttan, Mitesh M Khapra, Raj Dabre

RomanLens: The Role Of Latent Romanization In Multilinguality In LLMs (2025.findings-acl)

Challenge:	Existing datasets that cover only a fraction of Indian languages lack the breadth needed to generalize beyond curated benchmarks.
Approach:	They propose to build the largest speech translation dataset for Indian languages . they use a three-step methodology to gather data and train a model that performs better .
Outcome:	The proposed model improves on existing models and is open-source with permissive licenses.

Copied to clipboard

Alan Saji, Jaavid Aktar Husain, Thanmay Jayakumar, Raj Dabre, Anoop Kunchukuttan, Ratish Puduppully

RiddleBench: A New Generative Reasoning Benchmark for LLMs (2026.findings-eacl)

Challenge:	Large Language Models (LLMs) exhibit strong multilingual performance despite training on English-centric corpora.
Approach:	They propose to use Romanization as a potential bridge in multilingual processing . they propose to encode semantic concepts similarly across native and Romanized scripts .
Outcome:	The proposed model encodes semantic concepts across native and Romanized scripts, suggesting a shared underlying representation.

Copied to clipboard

Deepon Halder, Alan Saji, Thanmay Jayakumar, Anoop Kunchukuttan, Ratish Puduppully, Raj Dabre

Reasoning

Challenge:	Large Language Models (LLMs) show remarkable capabilities, but complex reasoning skills require deeper investigation.
Approach:	They propose a benchmark of 1,737 puzzles to test reasoning beyond simple pattern matching.
Outcome:	The proposed model performs poorly when faced with reordered constraints or irrelevant information.

CTQScorer: Combining Multiple Features for In-context Example Selection for Machine Translation (2023.findings-emnlp)

Copied to clipboard

Aswanth M, Ratish Puduppully, Raj Dabre, Anoop Kunchukuttan

Harnessing Cross-lingual Features to Improve Cognate Detection for Low-resource Languages (2020.coling-main)

Challenge:	Large language models have demonstrated the capability to perform on machine translation when the input is prompted with a few examples.
Approach:	They propose a regression model that combine features influencing example selection to maximize translation quality.
Outcome:	The proposed model outperforms random selection and strong single-factor baselines on multiple language pairs and language models.

Copied to clipboard

Diptesh Kanojia, Raj Dabre, Shubham Dewangan, Pushpak Bhattacharyya, Gholamreza Haffari, Malhar Kulkarni

Overview of the 6th Workshop on Asian Translation (D19-52)

Challenge:	a study of 14 Indian languages shows that cognates can be detected by word embeddings . cognates are variants of the same lexical form across languages .
Approach:	They propose to use cross-lingual word embeddings to detect cognates among 14 Indian languages . they then evaluate the impact of their method on neural machine translation .
Outcome:	The proposed method improves on a dataset of 12 Indian languages . it also improves quality of the extracted cognates by up to 2.76 BLEU .

Copied to clipboard

Toshiaki Nakazawa, Nobushige Doi, Shohei Higashiyama, Chenchen Ding, Raj Dabre, Hideya Mino, Isao Goto, Win Pa Pa, Anoop Kunchukuttan, Yusuke Oda, Shantipriya Parida, Ondřej Bojar, Sadao Kurohashi

Kreyòl-MT: Building MT for Latin American, Caribbean and Colonial African Creole Languages (2024.naacl-long)

Challenge:	The 6th workshop on Asian translation (WAT2019) was held in hong kong, hongkong, and hong kong.
Approach:	They present the results of the shared tasks from the 6th workshop on Asian translation (WAT2019) 25 teams participated in the shared task and 10 research paper submissions were accepted .
Outcome:	The results of the 6th workshop on Asian translation (WAT2019) include JaEn, JaZh scientific paper translation subtasks, Ja'En, ja'Ko, Ja’En patent translation sub tasks, Hi'En and My'En patent subtask and Ru'Ja news commentary translation task.

Copied to clipboard

Nathaniel Robinson, Raj Dabre, Ammon Shurtz, Rasul Dent, Onenamiyi Onesi, Claire Monroc, Loïc Grobol, Hasan Muhammad, Ashi Garg, Naome Etori, Vijay Murari Tiyyala, Olanrewaju Samuel, Matthew Stutzman, Bismarck Odoom, Sanjeev Khudanpur, Stephen Richardson, Kenton Murray

PUB: A Pragmatics Understanding Benchmark for Assessing LLMs’ Pragmatics Capabilities (2024.findings-acl)

Challenge:	Creole languages are used in much of Latin America, Africa and the Caribbean . a large multilingual bitext like ours has potential to build the best yet or first ever MT models for many languages .
Approach:	They present the largest cumulative dataset to date for Creole language MT . they provide MT models supporting all 41 Creoles in 172 translation directions .
Outcome:	The proposed model outperforms a genre-specific Creole MT model on its own benchmark for 23 of 34 translation directions.

Copied to clipboard

Settaluri Sravanthi, Meet Doshi, Pavan Tankala, Rudra Murthy, Raj Dabre, Pushpak Bhattacharyya

Semantic Text Processing

Challenge:	Pragmatics understanding is not well studied in LLMs, but their understanding of pragmatics is lacking.
Approach:	They propose to use a dataset to measure LLMs' understanding of pragmatics to evaluate their models.
Outcome:	The proposed dataset includes 14 tasks in four pragmatics phenomena, namely; Implicature, Presupposition, Reference, and Deixis.

IndicBART: A Pre-trained Model for Indic Natural Language Generation (2022.findings-acl)

Copied to clipboard

Raj Dabre, Himani Shrotriya, Anoop Kunchukuttan, Ratish Puduppully, Mitesh Khapra, Pratyush Kumar

When do Contrastive Word Alignments Improve Many-to-many Neural Machine Translation? (2022.findings-naacl)

Challenge:	IndicBART is a multilingual, sequence-to-sequence pre-trained model focusing on 11 Indic languages and English.
Approach:	They present a multilingual sequence-to-sequence pre-trained model for Indic languages . they evaluate it on two NLG tasks: Neural Machine Translation and extreme summarization .
Outcome:	The proposed model performs well on low-resource translation scenarios . Script sharing, multilingual training, and better utilization contribute to the performance.

Copied to clipboard

Zhuoyuan Mao, Chenhui Chu, Raj Dabre, Haiyue Song, Zhen Wan, Sadao Kurohashi

IndicMT Eval: A Dataset to Meta-Evaluate Machine Translation Metrics for Indian Languages (2023.acl-long)

Challenge:	Existing methods to improve pre-training for many-to-many neural machine translation use manual cleaning of bilingual dictionaries, which are unavailable for most language pairs.
Approach:	They propose a word-level contrastive objective to leverage word alignments for many-to-many neural machine translation (NMT) Empirical results show that this leads to 0.8 BLEU gains for several language pairs.
Outcome:	Empirical results show that the proposed objective leads to 0.8 BLEU gains for several language pairs.

Copied to clipboard

Ananya Sai B, Tanay Dixit, Vignesh Nagarajan, Anoop Kunchukuttan, Pratyush Kumar, Mitesh M. Khapra, Raj Dabre

Cross-Lingual Auto Evaluation for Assessing Multilingual LLMs (2025.acl-long)

Challenge:	Recent studies on machine translation systems focus on high-resource languages, but focus has shifted to low-resourced languages.
Approach:	They evaluate 16 metrics from a multidimensional quality metric dataset . they show pre-trained metrics have higher correlations with annotator scores .
Outcome:	The proposed evaluations show that pre-trained metrics outperform COMET on Indian languages.

Copied to clipboard

Sumanth Doddapaneni, Mohammed Safi Ur Rahman Khan, Dilip Venkatesh, Raj Dabre, Anoop Kunchukuttan, Mitesh M Khapra

Limited-Resource Adapters Are Regularizers, Not Linguists (2025.acl-short)

Challenge:	Evaluating machine-generated text remains a challenge in NLP for non-English languages . current evaluation frameworks focus on English, revealing a gap in multilingual evaluations .
Approach:	They propose a cross-lingual auto evaluation framework that includes evaluator LLMs and a test set specifically designed for multilingual evaluation.
Outcome:	The proposed model aligns more closely with human judgments than proprietary models on non-English language evaluations.

Copied to clipboard

Marcell Fekete, Nathaniel Romney Robinson, Ernests Lavrinovics, Djeride Jean-Baptiste, Raj Dabre, Johannes Bjerva, Heather Lent

Responsible & Trustworthy NLP

Challenge:	Existing studies show that cross-lingual transfer from high-resource languages is promising for low-resourced machine translation.
Approach:	They propose to use adapter souping and cross-attention fine-tuning to leverage language transfer for Creoles, an under-served group of low-resource languages.
Outcome:	The proposed method improves performance over baselines but not meaningfully with adapters.

NGLUEni: Benchmarking and Adapting Pretrained Language Models for Nguni Languages (2024.lrec-main)

Copied to clipboard

Francois Meyer, Haiyue Song, Abhisek Chakrabarty, Jan Buys, Raj Dabre, Hideki Tanaka

Exploring the Impact of Layer Normalization for Zero-shot Neural Machine Translation (2023.acl-short)

Challenge:	Nguni languages have over 20 million home language speakers in South Africa . there has been considerable growth in the datasets for these languages, but no analysis of the performance of NLP models for these language has been reported across languages and tasks.
Approach:	They compile publicly available datasets for natural language understanding and generation, spanning 6 tasks and 11 datasets.
Outcome:	The proposed models outperform existing models and large-scale adapted models on cross-lingual transfer and machine translation.

Copied to clipboard

Zhuoyuan Mao, Raj Dabre, Qianying Liu, Haiyue Song, Chenhui Chu, Sadao Kurohashi

BERTSeg: BERT Based Unsupervised Subword Segmentation for Neural Machine Translation (2022.aacl-short)

Challenge:	Recent studies have shown that layer normalization (LayerNorm) overfits training data and therefore has low generalizability for ZST.
Approach:	They propose to use the Transformer architecture to set the default layer normalization setting for zero-shot translation (ZST) they also propose to set LayerNorm after residual connections to outperform PreNorm by 12.3 BLEU points.
Outcome:	The proposed model outperforms the current model by 12.3 BLEU points on 54 directions on OPUS, IWSLT, and Europarl datasets.

Copied to clipboard

Haiyue Song, Raj Dabre, Zhuoyuan Mao, Chenhui Chu, Sadao Kurohashi

Exploiting Multilingualism through Multistage Fine-Tuning for Low-Resource Neural Machine Translation (D19-1)

Challenge:	Existing subword segmenters are frequency-based without semantics information or neural-based but trained on parallel corpora.
Approach:	They propose an unsupervised neural subword segmenter for neural machine translation that utilizes contextualized semantic embeddings of words from characterBERT and maximizes the generation probability of subword segments.
Outcome:	The proposed method improves translation performance on ALT, IWSLT15 Vi->En, WMT16 Ro->En and WMT15 Fi->En datasets.

Copied to clipboard

Raj Dabre, Atsushi Fujita, Chenhui Chu

DecoMT: Decomposed Prompting for Machine Translation Between Related Languages using Large Language Models (2023.emnlp-main)

Challenge:	Using multi-parallel corpora for transfer learning is a useful technique for low-resource NMT.
Approach:	They compare multi-parallel corpora for transfer learning in a low-resource setting . their results show that multi-paralleled corpors are extremely useful .
Outcome:	The proposed model can give 3–9 BLEU score gains over a one-to-one model.

Copied to clipboard

Ratish Puduppully, Anoop Kunchukuttan, Raj Dabre, Ai Ti Aw, Nancy Chen