Open Research Infrastructure
Collection
Open, link-verified tools and catalogs for finding and evaluating research — each citable with a DOI — plus the papers behind them. • 12 items • Updated
task stringclasses 11
values | id stringlengths 3 24 | name stringlengths 4 45 | size stringlengths 7 29 | sizeN float64 22 8.6M | license stringlengths 3 25 | year int64 2.01k 2.03k | source stringlengths 12 50 | link stringlengths 23 127 | verified timestamp[s]date 2026-01-30 00:00:00 2026-09-08 00:00:00 | desc stringlengths 34 251 | bibtex stringlengths 231 11.1k ⌀ |
|---|---|---|---|---|---|---|---|---|---|---|---|
hate | bdshs | BD-SHS | 50,281 comments | 50,281 | CC BY-NC 4.0 | 2,022 | BD-SHS (LREC 2022) | https://github.com/naurosromim/hate-speech-dataset-for-Bengali-social-media | 2026-06-05T00:00:00 | Hierarchically annotated hate speech: target, type, severity. | @inproceedings{romim-etal-2022-bd,
title = "{BD}-{SHS}: A Benchmark Dataset for Learning to Detect Online {B}angla Hate Speech in Different Social Contexts",
author = "Romim, Nauros and
Ahmed, Mosahed and
Islam, Md Saiful and
Sen Sharma, Arnab and
Talukder, Hriteshwar and
Amin... |
hate | bhs | Bengali Hate Speech | 3,418 labelled statements | 3,418 | MIT | 2,020 | Bengali Hate Speech v1.0 (Karim et al., 2020) | https://github.com/rezacsedu/Bengali-Hate-Speech-Dataset | 2026-09-08T00:00:00 | v1.0 release with one label per statement: geopolitical, religious, personal, gender abusive, or political hate. Repository license: MIT. | @misc{karim2020classificationbenchmarksunderresourcedbengali,
title={Classification Benchmarks for Under-resourced Bengali Language based on Multichannel Convolutional-LSTM Network},
author={Md. Rezaul Karim and Bharathi Raja Chakravarthi and John P. McCrae and Michael Cochez},
year={2020},
epri... |
hate | tbolid | TB-OLID | 5,000 comments | 5,000 | CC BY 4.0 | 2,023 | Transliterated offensive language ID | https://github.com/LanguageTechnologyLab/TB-OLID | 2026-03-08T00:00:00 | Offensive language in transliterated (Banglish) text. | @inproceedings{raihan-etal-2023-offensive,
title = "Offensive Language Identification in Transliterated and Code-Mixed {B}angla",
author = "Raihan, Md Nishat and
Tanmoy, Umma and
Islam, Anika Binte and
North, Kai and
Ranasinghe, Tharindu and
Anastasopoulos, Antonios and
... |
hate | offmix3l | OffMix-3L | 1,001 instances | 1,001 | AGPL-3.0 | 2,023 | OffMix-3L (SocialNLP 2023) | https://github.com/LanguageTechnologyLab/OffMix-3L | 2026-07-19T00:00:00 | Bangla-English-Hindi code-mixed offensive language identification test set. | @inproceedings{goswami-etal-2023-offmix,
title = "{O}ff{M}ix-3{L}: A Novel Code-Mixed Test Dataset in {B}angla-{E}nglish-{H}indi for Offensive Language Identification",
author = "Goswami, Dhiman and
Raihan, Md Nishat and
Mahmud, Antara and
Anastasopoulos, Antonios and
Zampieri, Marco... |
hate | banth | BanTH | 37,300 samples | 37,300 | MIT | 2,025 | BanTH (Findings NAACL 2025) | https://github.com/farhanishmam/BanTH | 2026-07-19T00:00:00 | Multi-label transliterated (romanized) Bangla hate speech dataset from YouTube across 7 target classes. | @inproceedings{haider-etal-2025-banth,
title = "{B}an{TH}: A Multi-label Hate Speech Detection Dataset for Transliterated {B}angla",
author = "Haider, Fabiha and
Shifat, Fariha Tanjim and
Ishmam, Md Farhan and
Sourove, Md Sakib Ul Rahman and
Barua, Deeparghya Dutta and
Fahim, ... |
hate | alert | ALERT | 4,027 texts | 4,027 | CC BY 4.0 | 2,025 | ALERT (Data in Brief 2025) | https://data.mendeley.com/datasets/f4xz5d4fzd/1 | 2026-07-26T00:00:00 | Bangla religious-aggression texts with English translations, labelled hate speech, vandalism, atrocity, or no aggression. | @article{Rashid_2025, title={ALERT: A benchmark Bengali dataset for identifying and categorizing religiously aggressive texts}, volume={63}, ISSN={2352-3409}, url={http://dx.doi.org/10.1016/j.dib.2025.112094}, DOI={10.1016/j.dib.2025.112094}, journal={Data in Brief}, publisher={Elsevier BV}, author={Rashid, Suhana Bint... |
hate | banglatoco | Bangla-ToCo | 1,004 comments | 1,004 | CC BY 4.0 | 2,025 | Bangla-ToCo (Data in Brief 2025) | https://data.mendeley.com/datasets/gphbs7vsbz/3 | 2026-07-26T00:00:00 | Context-aware toxic-comment corpus retaining news metadata and neighbouring comments, with balanced toxic/non-toxic labels. | @article{Rupa_2025, title={Bangla-ToCo: A context-aware dataset for Bengali toxic comment detection}, volume={63}, ISSN={2352-3409}, url={http://dx.doi.org/10.1016/j.dib.2025.112277}, DOI={10.1016/j.dib.2025.112277}, journal={Data in Brief}, publisher={Elsevier BV}, author={Rupa, Sayma Akter and Anwar, Md. Musfique and... |
hate | banhate | BanHate | 19,203 comments | 19,203 | MIT | 2,025 | BanHate (BLP 2025) | https://huggingface.co/datasets/aplycaebous/BanHate | 2026-08-23T00:00:00 | YouTube comments annotated for binary hate, hate-category, and target-group labels. | @inproceedings{raquib-etal-2025-banhate,
title = "{B}an{H}ate: An Up-to-Date and Fine-Grained {B}angla Hate Speech Dataset",
author = "Raquib, Faisal Hossain and
Mazumder, Akm Moshiur Rahman and
Fuad, Md Tahmid Hasan and
Ishmam, Md Farhan and
Fahim, Md",
editor = "Alam, Firoj and
Kar, Sud... |
llm | benqa | BEnQA | 5,161 questions | 5,161 | CC BY-NC-SA 4.0 | 2,024 | BEnQA (2024) | https://github.com/sheikhshafayat/BEnQA | 2026-07-19T00:00:00 | Parallel Bangla–English science exam questions for LLM evaluation. | @article{shafayat2024benqa,
title = {{BE}n{QA}: A Question Answering and Reasoning Benchmark for {B}engali and {E}nglish},
author = {Shafayat, Sheikh and others},
journal = {arXiv preprint arXiv:2403.10900},
year = {2024}
} |
llm | globalmmlu | Global-MMLU (bn) | 14,327 questions | 14,327 | Apache 2.0 | 2,024 | Global-MMLU (2024) | https://huggingface.co/datasets/CohereLabs/Global-MMLU | 2026-07-19T00:00:00 | Culturally adapted MMLU; Bengali split. | @misc{singh2025globalmmluunderstandingaddressing,
title={Global MMLU: Understanding and Addressing Cultural and Linguistic Biases in Multilingual Evaluation},
author={Shivalika Singh and Angelika Romanou and Clémentine Fourrier and David I. Adelani and Jian Gang Ngui and Daniel Vila-Suero and Peerat Limkonc... |
llm | megaverse | MEGAVERSE (bn subset) | 22 datasets | 22 | Mixed | 2,024 | MEGAVERSE (NAACL 2024) | https://github.com/microsoft/MEGAVERSE | 2026-08-23T00:00:00 | Multilingual LLM evaluation suite including Bangla tasks. | @inproceedings{ahuja-etal-2024-megaverse,
title = "{MEGAVERSE}: Benchmarking Large Language Models Across Languages, Modalities, Models and Tasks",
author = "Ahuja, Sanchit and
Aggarwal, Divyanshu and
Gumma, Varun and
Watts, Ishaan and
Sathe, Ashutosh and
Ochieng, Millicent a... |
llm | breaso | B-REASO | 13,497 questions | 13,497 | MIT | 2,025 | B-REASO (Findings EMNLP 2025) | https://huggingface.co/datasets/krarit/b-reaso | 2026-07-19T00:00:00 | Bengali multiple-choice evaluation suite over 50 subjects at four difficulty levels, plus the harder B-REASO HEAVY subset. | @inproceedings{hosain-morol-2025-b,
title = {{B}-{REASO}: A Multi-Level Multi-Faceted {B}engali Evaluation Suite for Foundation Models},
author = {Hosain, Md Tanzib and Morol, Md Kishor},
booktitle = {Findings of the Association for Computational Linguistics: EMNLP 2025},
year = {2025},
pages = {9260--9274},
... |
llm | bnmmlu | BnMMLU | 134,375 question-option pairs | 134,375 | CC BY-SA 4.0 | 2,025 | BnMMLU (2025) | https://github.com/samanjoy2/bnmmlu | 2026-07-19T00:00:00 | Bengali massive multitask language understanding benchmark spanning 23 domains. | @inproceedings{joy-shatabda-2026-bnmmlu,
title = "{B}n{MMLU}: Measuring Massive Multitask Language Understanding in {B}engali",
author = "Joy, Saman Sarker and
Shatabda, Swakkhar",
editor = "Liakata, Maria and
Moreira, Viviane P. and
Zhang, Jiajun and
Jurgens, David",
booktit... |
llm | banglamath | BanglaMATH | ~1,700 problems | 1,700 | Research only | 2,025 | BanglaMATH (MathNLP 2025) | https://github.com/TabiaTanzin/BanglaMATH-A-Bangla-benchmark-dataset-for-testing-LLM-mathematical-reasoning-at-grades-6-7-and-8 | 2026-07-19T00:00:00 | Bangla grade 6-8 math word problems for testing LLM mathematical reasoning. | @inproceedings{prama-etal-2025-banglamath,
title = "{B}angla{MATH} : A {B}angla benchmark dataset for testing {LLM} mathematical reasoning at grades 6, 7, and 8",
author = "Prama, Tabia Tanzin and
Danforth, Christopher M. and
Dodds, Peter",
editor = "Valentino, Marco and
Ferreira, Debor... |
llm | somadhan | SOMADHAN | 8,792 problems | 8,792 | CC BY 4.0 | 2,025 | SOMADHAN (2025) | https://data.mendeley.com/datasets/34bs5cxk9j/2 | 2026-07-26T00:00:00 | Bengali math word problems with step-by-step solutions adapted from the GSM8K train and test sets. | @misc{paul2025leveraginglargelanguagemodels,
title={Leveraging Large Language Models for Bengali Math Word Problem Solving with Chain of Thought Reasoning},
author={Bidyarthi Paul and Jalisha Jashim Era and Mirazur Rahman Zim and Tahmid Sattar Aothoi and Faisal Muhammad Shah},
year={2025},
eprin... |
llm | banglasocialbench | BanglaSocialBench | 1,719 instances | 1,719 | CC BY 4.0 | 2,026 | BanglaSocialBench (ACL SRW 2026) | https://huggingface.co/datasets/sijantanvir/BanglaSocialBench | 2026-08-23T00:00:00 | Sociopragmatic LLM benchmark over Bangla address terms, kinship reasoning, and social customs. | @inproceedings{sijan-etal-2026-banglasocialbench,
title = "{B}angla{S}ocial{B}ench: A Benchmark for Evaluating Sociopragmatic and Cultural Alignment of {LLM}s in Bangladeshi Social Interaction",
author = "Sijan, Tanvir Ahmed and
Rifat, S. M. Golam and
Partha, Pankaj and
Islam, Md. Tanjeed and
An... |
mt | banglanmt | BanglaNMT Parallel Corpus | 2.75M pairs | 2,750,000 | Research only | 2,020 | Not Low-Resource Anymore (EMNLP 2020) | https://github.com/csebuetnlp/banglanmt | 2026-06-28T00:00:00 | High-quality Bangla–English parallel corpus built with aligner ensembling. | @inproceedings{hasan-etal-2020-low,
title = {Not Low-Resource Anymore: Aligner Ensembling, Batch Filtering, and New Datasets for {B}engali-{E}nglish Machine Translation},
author = {Hasan, Tahmid and Bhattacharjee, Abhik and others},
booktitle = {EMNLP},
year = {2020}
} |
mt | flores | FLORES-200 (bn) | 2,009 public sentences | 2,009 | CC BY-SA 4.0 | 2,022 | No Language Left Behind (2022) | https://github.com/facebookresearch/flores | 2026-09-08T00:00:00 | Bengali ben_Beng evaluation data with 997 dev and 1,012 devtest sentences; excludes the hidden test split. | @misc{nllbteam2022languageleftbehindscaling,
title={No Language Left Behind: Scaling Human-Centered Machine Translation},
author={NLLB Team and Marta R. Costa-jussà and James Cross and Onur Çelebi and Maha Elbayad and Kenneth Heafield and Kevin Heffernan and Elahe Kalbassi and Janice Lam and Daniel Licht an... |
mt | banglaparaphrase | BanglaParaphrase | 466,630 pairs | 466,630 | CC BY-NC-SA 4.0 | 2,022 | BanglaParaphrase (AACL 2022) | https://huggingface.co/datasets/csebuetnlp/BanglaParaphrase | 2026-07-19T00:00:00 | High-quality synthetic paraphrase pairs. | @inproceedings{akil-etal-2022-banglaparaphrase,
title = "{B}angla{P}araphrase: A High-Quality {B}angla Paraphrase Dataset",
author = "Akil, Ajwad and
Sultana, Najrin and
Bhattacharjee, Abhik and
Shahriyar, Rifat",
editor = "He, Yulan and
Ji, Heng and
Li, Sujian and
... |
mt | samanantar | Samanantar (bn–en) | 8.6M pairs | 8,600,000 | CC0 | 2,022 | Samanantar (TACL 2022) | https://indicnlp.ai4bharat.org/samanantar/ | 2026-07-19T00:00:00 | Largest publicly available bn–en parallel collection. | @article{ramesh-etal-2022-samanantar,
title = "Samanantar: The Largest Publicly Available Parallel Corpora Collection for 11 {I}ndic Languages",
author = "Ramesh, Gowtham and
Doddapaneni, Sumanth and
Bheemaraj, Aravinth and
Jobanputra, Mayank and
AK, Raghavan and
Sharma, Ajite... |
mt | banglatlit | BanglaTLit | 42,705 pairs | 42,705 | MIT | 2,024 | BanglaTLit (Findings EMNLP 2024) | https://github.com/farhanishmam/BanglaTLit | 2026-07-19T00:00:00 | Benchmark for back-transliteration of romanized Bangla to Bangla script. | @inproceedings{fahim-etal-2024-banglatlit,
title = "{B}angla{TL}it: A Benchmark Dataset for Back-Transliteration of {R}omanized {B}angla",
author = "Fahim, Md and
Shifat, Fariha Tanjim and
Haider, Fabiha and
Barua, Deeparghya Dutta and
Sourove, MD Sakib Ul Rahman and
Ishmam, M... |
mt | vashantor | Vashantor | 32,500 sentences | 32,500 | CC BY 4.0 | 2,023 | Vashantor (2023) | https://data.mendeley.com/datasets/bj5jgk878b/2 | 2026-07-19T00:00:00 | Translation benchmark from five Bangla regional dialects to standard Bangla. | @misc{faria2025vashantorlargescalemultilingualbenchmark,
title={Vashantor: A Large-scale Multilingual Benchmark Dataset for Automated Translation of Bangla Regional Dialects to Bangla Language},
author={Fatema Tuj Johora Faria and Mukaffi Bin Moin and Ahmed Al Wase and Mehidi Ahmmed and Md. Rabius Sani and ... |
mt | banglaregionaltextcorpus | BanglaRegionalTextCorpus | 4,653 sentences | 4,653 | CC BY 4.0 | 2,026 | BanglaRegionalTextCorpus (Data in Brief 2026) | https://data.mendeley.com/datasets/92r62h4k5k/4 | 2026-07-26T00:00:00 | Regional sentences from Rangpur, Barisal, Narail, and Khulna with standard Bangla and English translations. | @article{Ahmed_2026, title={BanglaRegionalTextCorpus: A curated dataset for four regional bangla dialects with standard Bangla and English translation}, volume={65}, ISSN={2352-3409}, url={http://dx.doi.org/10.1016/j.dib.2026.112585}, DOI={10.1016/j.dib.2026.112585}, journal={Data in Brief}, publisher={Elsevier BV}, au... |
mt | kothon | Kothon | 17,300 sentence pairs | 17,300 | CC BY 4.0 | 2,026 | Kothon (Data in Brief 2026) | https://data.mendeley.com/datasets/2fv6vf9v2z/4 | 2026-07-26T00:00:00 | Parallel Chittagonian-Bangla and Sylheti-Bangla corpora with 8,000 and 9,300 sentence pairs. | @article{Faisal_2026, title={Kothon: A large-scale dataset for machine translation of the Chittagonian and Sylheti dialects into standard Bangla}, volume={66}, ISSN={2352-3409}, url={http://dx.doi.org/10.1016/j.dib.2026.112789}, DOI={10.1016/j.dib.2026.112789}, journal={Data in Brief}, publisher={Elsevier BV}, author={... |
ner | multiconer | MultiCoNER (bn) | 15,300 sentences | 15,300 | CC BY 4.0 | 2,022 | SemEval-2022 Task 11 | https://registry.opendata.aws/multiconer/ | 2026-06-10T00:00:00 | Complex NER: creative works, groups, products; Bangla track. | @inproceedings{malmasi-etal-2022-semeval,
title = "{S}em{E}val-2022 Task 11: Multilingual Complex Named Entity Recognition ({M}ulti{C}o{NER})",
author = "Malmasi, Shervin and
Fang, Anjie and
Fetahu, Besnik and
Kar, Sudipta and
Rokhlenko, Oleg",
editor = "Emerson, Guy and
S... |
ner | bner | B-NER | 22,144 sentences | 22,144 | MIT | 2,023 | B-NER (IEEE Access 2023) | https://www.kaggle.com/datasets/mdzahidulhaquealvi/b-ner | 2026-07-19T00:00:00 | Largest gold-standard Bangla NER corpus, 8 entity types. | @article{Haque_2023, title={B-NER: A Novel Bangla Named Entity Recognition Dataset With Largest Entities and Its Baseline Evaluation}, volume={11}, ISSN={2169-3536}, url={http://dx.doi.org/10.1109/ACCESS.2023.3267746}, DOI={10.1109/access.2023.3267746}, journal={IEEE Access}, publisher={Institute of Electrical and Elec... |
ner | wikiann | WikiANN (bn) | 12,000 sentences | 12,000 | ODC-By; research use only | 2,019 | Massively Multilingual Transfer for NER (ACL 2019) | https://huggingface.co/datasets/unimelb-nlp/wikiann | 2026-09-08T00:00:00 | Silver standard Wikipedia PER/LOC/ORG annotations in the Rahimi et al. balanced splits; Bengali train/validation/test contain 10,000/1,000/1,000 sentences. The license field records the original release's ODC-By and 'For research use only' statements. | @inproceedings{rahimi-etal-2019-massively,
title = "Massively Multilingual Transfer for {NER}",
author = "Rahimi, Afshin and
Li, Yuan and
Cohn, Trevor",
editor = "Korhonen, Anna and
Traum, David and
M{\`a}rquez, Llu{\'i}s",
booktitle = "Proceedings of the 57th Annual Meeting ... |
ner | ancholikner | ANCHOLIK-NER | 17,405 sentences | 17,405 | CC BY 4.0 | 2,025 | ANCHOLIK-NER (2025) | https://data.mendeley.com/datasets/gbkszkt8z3/4 | 2026-07-19T00:00:00 | Bangla regional NER across five dialects with 10 entity types, BIO-tagged. | @misc{paul2025ancholiknerbenchmarkdatasetbangla,
title={ANCHOLIK-NER: A Benchmark Dataset for Bangla Regional Named Entity Recognition},
author={Bidyarthi Paul and Faika Fairuj Preotee and Shuvashis Sarker and Shamim Rahim Refat and Shifat Islam and Tashreef Muhammad and Mohammad Ashraful Hoque and Shahriar... |
ner | banglameder | Bangla-MedER | 2,980 texts | 2,980 | CC BY 4.0 | 2,025 | Bangla-MedER (Data in Brief 2026) | https://data.mendeley.com/datasets/jt4gywvwtj/2 | 2026-07-26T00:00:00 | Manually annotated medical text covering six entity types, with a separate English translation. | @article{Sheikh_2026, title={Bangla-MedER: An annotated Bangla dataset for multi-type medical entity recognition from medical text}, volume={66}, ISSN={2352-3409}, url={http://dx.doi.org/10.1016/j.dib.2026.112705}, DOI={10.1016/j.dib.2026.112705}, journal={Data in Brief}, publisher={Elsevier BV}, author={Sheikh, Rubel ... |
ner | bannerd | BanNERD | 85,175 sentences | 85,175 | CC BY-NC-ND 4.0 | 2,025 | BanNERD (Findings NAACL 2025) | https://github.com/eblict-sd07/BanNERD | 2026-08-23T00:00:00 | Human-annotated Bangla NER corpus spanning 29 domains and 10 entity classes. | @inproceedings{mahtab-etal-2025-bannerd,
title = "{B}an{NERD}: A Benchmark Dataset and Context-Driven Approach for {B}angla Named Entity Recognition",
author = "Mahtab, Md. Motahar and
Khan, Faisal Ahamed and
Islam, Md. Ekramul and
Chowdhury, Md. Shahad Mahmud and
Chowdhury, Labib Imam and
... |
pos | udbru | UD Bengali-BRU Treebank | 320 tokens / 56 sentences | 56 | CC BY-SA 4.0 | 2,021 | Universal Dependencies (since v2.9) | https://universaldependencies.org/treebanks/bn_bru/ | 2026-08-23T00:00:00 | Tiny Universal Dependencies sample treebank with UPOS and dependency relations. Citation covers the UD 2.9 release, including this treebank. | @misc{11234/1-4611,
title = {Universal Dependencies 2.9},
author = {Zeman, Daniel and Nivre, Joakim and Abrams, Mitchell and Ackermann, Elia and Aepli, No{\"e}mi and Aghaei, Hamid and Agi{\'c}, {\v Z}eljko and Ahmadi, Amir and Ahrenberg, Lars and Ajede, Chika Kennedy and Aleksandravi{\v c}i{\=u}t{\.e}, Ga... |
pos | snltr | SNLTR POS Corpus | ~7,390 sentences | 7,390 | Research only | 2,010 | Society for Natural Language Technology Research | http://nltr.org/snltr-software/ | 2026-07-20T00:00:00 | Classic manually tagged corpus, 40-tag tagset. | null |
qa | squadbn | SQuAD_bn | 132,777 QA pairs | 132,777 | CC BY-NC-SA 4.0 | 2,021 | BanglaBERT benchmarks | https://huggingface.co/datasets/csebuetnlp/squad_bn | 2026-07-19T00:00:00 | Machine-translated + human-curated SQuAD 2.0 for Bangla. | @inproceedings{bhattacharjee-etal-2022-banglabert,
title = "{B}angla{BERT}: Language Model Pretraining and Benchmarks for Low-Resource Language Understanding Evaluation in {B}angla",
author = "Bhattacharjee, Abhik and
Hasan, Tahmid and
Ahmad, Wasi and
Mubasshir, Kazi Samin and
Islam,... |
qa | banglarqa | BanglaRQA | 14,889 QA pairs | 14,889 | CC BY-NC 4.0 | 2,022 | BanglaRQA (Findings EMNLP 2022) | https://github.com/sartajekram419/BanglaRQA | 2026-06-15T00:00:00 | Human-annotated reading comprehension with answerable/unanswerable questions. | @inproceedings{ekram-etal-2022-banglarqa,
title = "{B}angla{RQA}: A Benchmark Dataset for Under-resourced {B}angla Language Reading Comprehension-based Question Answering with Diverse Question-Answer Types",
author = "Ekram, Syed Mohammed Sartaj and
Rahman, Adham Arik and
Altaf, Md. Sajid and
... |
qa | tydiqa | TyDi QA (bn) | 11,096 questions | 11,096 | Apache 2.0 | 2,020 | TyDi QA primary tasks v1.0 (TACL 2020) | https://github.com/google-research-datasets/tydiqa | 2026-09-08T00:00:00 | Bengali primary-task data with 10,768 train and 328 dev questions for passage selection and minimal answer extraction; not the GoldP subset. | @article{clark-etal-2020-tydi,
title = "{T}y{D}i {QA}: A Benchmark for Information-Seeking Question Answering in Typologically Diverse Languages",
author = "Clark, Jonathan H. and
Choi, Eunsol and
Collins, Michael and
Garrette, Dan and
Kwiatkowski, Tom and
Nikolaev, Vitaly an... |
qa | nctbqa | NCTB-QA | 87,805 QA pairs | 87,805 | CC BY-NC 4.0 | 2,026 | NCTB-QA (arXiv 2026) | https://huggingface.co/datasets/eyasir2047/NCTB-QA | 2026-08-23T00:00:00 | Educational QA from 50 NCTB textbooks with answerable, unanswerable, and adversarial instances. | @misc{eyasir2026nctbqalargescalebanglaeducational,
title={NCTB-QA: A Large-Scale Bangla Educational Question Answering Dataset and Benchmarking Performance},
author={Abrar Eyasir and Tahsin Ahmed and Muhammad Ibrahim},
year={2026},
eprint={2603.05462},
archivePrefix={arXiv},
primaryClass={cs.CL},
url={htt... |
sentiment | sentnob | SentNoB | 15,728 comments | 15,728 | CC BY-ND 4.0 | 2,021 | SentNoB (Findings EMNLP 2021) | https://github.com/KhondokerIslam/SentNoB | 2026-09-08T00:00:00 | Sentiment on noisy user comments from news and social platforms across 13 domains. | @inproceedings{islam-etal-2021-sentnob-dataset,
title = "{S}ent{N}o{B}: A Dataset for Analysing Sentiment on Noisy {B}angla Texts",
author = "Islam, Khondoker Ittehadul and
Kar, Sudipta and
Islam, Md Saiful and
Amin, Mohammad Ruhul",
editor = "Moens, Marie-Francine and
Huang, Xua... |
sentiment | bemoc | BEmoC | 7,000 texts | 7,000 | Research only | 2,022 | Emotion classification corpus | https://github.com/avishek-018/BEmoC | 2026-07-19T00:00:00 | Six-class emotion corpus (anger, disgust, fear, joy, sadness, surprise). | @Article{Iqbal2022,
author="Iqbal, MD. Asif
and Das, Avishek
and Sharif, Omar
and Hoque, Mohammed Moshiul
and Sarker, Iqbal H.",
title="BEmoC: A Corpus for Identifying Emotion in Bengali Texts",
journal="SN Computer Science",
year="2022",
month="Jan",
day="17",
volume="3",
number="2",
pages="135",
abstract="Emotion cla... |
sentiment | absa | ABSA Cricket & Restaurant | 5,038 comments | 5,038 | Research only | 2,018 | Aspect-based sentiment (cricket, restaurant) | https://github.com/AtikRahman/Bangla_ABSA_Datasets | 2026-04-02T00:00:00 | Aspect-based sentiment in two domains. | @article{Rahman_2018, title={Datasets for Aspect-Based Sentiment Analysis in Bangla and Its Baseline Evaluation}, volume={3}, ISSN={2306-5729}, url={http://dx.doi.org/10.3390/data3020015}, DOI={10.3390/data3020015}, number={2}, journal={Data}, publisher={MDPI AG}, author={Rahman, Md. Atikur and Kumar Dey, Emon}, year={... |
sentiment | banglabook | BanglaBook | 158,065 reviews | 158,065 | CC BY-NC-SA 4.0 | 2,023 | BanglaBook (Findings ACL 2023) | https://github.com/mohsinulkabir14/BanglaBook | 2026-07-19T00:00:00 | Large-scale Bangla book-review sentiment dataset in three classes (positive, negative, neutral). | @inproceedings{kabir-etal-2023-banglabook,
title = "{B}angla{B}ook: A Large-scale {B}angla Dataset for Sentiment Analysis from Book Reviews",
author = "Kabir, Mohsinul and
Bin Mahfuz, Obayed and
Raiyan, Syed Rifat and
Mahmud, Hasan and
Hasan, Md Kamrul",
editor = "Rogers, Anna a... |
sentiment | motamot | Motamot | 7,058 instances | 7,058 | MIT | 2,024 | Motamot (2024) | https://github.com/mukaffi28/bengali-political-sentiment-analysis | 2026-07-19T00:00:00 | Bengali political sentiment dataset compiled from election-period news, labelled positive/negative. | @inproceedings{Johora_Faria_2024, title={Motamot: A Dataset for Revealing the Supremacy of Large Language Models Over Transformer Models in Bengali Political Sentiment Analysis}, url={http://dx.doi.org/10.1109/TENSYMP61132.2024.10752197}, DOI={10.1109/tensymp61132.2024.10752197}, booktitle={2024 IEEE Region 10 Symposiu... |
sentiment | sentmix3l | SentMix-3L | 1,007 instances | 1,007 | AGPL-3.0 | 2,023 | SentMix-3L (SEALP 2023) | https://github.com/LanguageTechnologyLab/SentMix-3L | 2026-07-19T00:00:00 | Gold-standard Bangla-English-Hindi code-mixed sentiment analysis test set. | @inproceedings{raihan-etal-2023-sentmix,
title = "{S}ent{M}ix-3{L}: A Novel Code-Mixed Test Dataset in {B}angla-{E}nglish-{H}indi for Sentiment Analysis",
author = "Raihan, Md Nishat and
Goswami, Dhiman and
Mahmud, Antara and
Anastasopoulos, Antonios and
Zampieri, Marcos",
editor... |
sentiment | emomix3l | EmoMix-3L | 1,071 instances | 1,071 | GPL-3.0 | 2,024 | EmoMix-3L (WILDRE 2024) | https://github.com/GoswamiDhiman/EmoMix-3L | 2026-07-19T00:00:00 | Bangla-English-Hindi code-mixed emotion detection test set. | @inproceedings{raihan-etal-2024-emomix,
title = "{E}mo{M}ix-3{L}: A Code-Mixed Dataset for {B}angla-{E}nglish-{H}indi for Emotion Detection",
author = "Raihan, Nishat and
Goswami, Dhiman and
Mahmud, Antara and
Anastasopoulos, Antonios and
Zampieri, Marcos",
editor = "Jha, Girish ... |
sentiment | bnsentmix | BnSentMix | 20,000 samples | 20,000 | MIT | 2,024 | BnSentMix (LoResLM 2025) | https://huggingface.co/datasets/aplycaebous/BnSentMix | 2026-07-19T00:00:00 | Bengali-English code-mixed sentiment dataset with 4 labels from social media and e-commerce. | @inproceedings{alam-etal-2025-bnsentmix,
title = "{B}n{S}ent{M}ix: A Diverse {B}engali-{E}nglish Code-Mixed Dataset for Sentiment Analysis",
author = "Alam, Sadia and
Ishmam, Md Farhan and
Alvee, Navid Hasin and
Siddique, Md Shahnewaz and
Hossain, Md Azam and
Kamal, Abu Raihan... |
sentiment | babsa | BABSA | 15,860 instances | 15,860 | CC BY-NC-SA 4.0 | 2,025 | BABSA (Data in Brief 2026) | https://data.mendeley.com/datasets/j7yb2sv263/1 | 2026-07-26T00:00:00 | Aspect-based sentiment corpus across 21 domains with manually annotated aspect terms and positive, neutral, or negative labels. | @article{Siam_2026, title={BABSA: A large scale bangla aspect based sentiment analysis dataset}, volume={65}, ISSN={2352-3409}, url={http://dx.doi.org/10.1016/j.dib.2026.112604}, DOI={10.1016/j.dib.2026.112604}, journal={Data in Brief}, publisher={Elsevier BV}, author={Siam, M.F.R. and Wafa, Narmeen}, year={2026}, mont... |
sentiment | banglasarc3 | BanglaSarc3 | 12,089 comments | 12,089 | CC BY 4.0 | 2,025 | BanglaSarc3 (Data in Brief 2025) | https://data.mendeley.com/datasets/7tn76wdhsr/1 | 2026-07-26T00:00:00 | Balanced ternary Facebook-comment corpus labelled sarcastic, non-sarcastic, or neutral. | @article{Biswas_2025, title={BanglaSarc3: A benchmark dataset for Bangla sarcasm detection from social media to advance Bangla NLP}, volume={62}, ISSN={2352-3409}, url={http://dx.doi.org/10.1016/j.dib.2025.111953}, DOI={10.1016/j.dib.2025.111953}, journal={Data in Brief}, publisher={Elsevier BV}, author={Biswas, Susmoy... |
ser | subesco | SUBESCO (SUST Bangla Emotional Speech Corpus) | 7,000 utterances / 7.7 hrs | 7,000 | CC BY 4.0 | 2,021 | Sultana et al. (SUST) / Zenodo | https://doi.org/10.5281/zenodo.4526477 | 2026-07-22T00:00:00 | Audio-only Bangla emotional speech; 20 actors (gender-balanced), 7 emotions. | @article{10.1371/journal.pone.0250173,
doi = {10.1371/journal.pone.0250173},
author = {Sultana, Sadia AND Rahman, M. Shahidur AND Selim, M. Reza AND Iqbal, M. Zafar},
journal = {PLOS ONE},
publisher = {Public Library of Science},
title = {SUST Bangla Emotional Speech Corpus (SUBESCO): An audio-only ... |
ser | banspemo | BANSpEmo | 792 utterances | 792 | CC BY 4.0 | 2,023 | Sultana et al. / Mendeley Data | https://data.mendeley.com/datasets/rdwn4bs5ky/2 | 2026-07-22T00:00:00 | Bangla emotional speech; 22 speakers (11M/11F), 6 emotions. | @article{Sultana_2025, title={BanSpEmo: a Bangla audio dataset for speech emotion recognition and its baseline evaluation}, volume={37}, ISSN={2502-4752}, url={http://dx.doi.org/10.11591/ijeecs.v37.i3.pp2044-2057}, DOI={10.11591/ijeecs.v37.i3.pp2044-2057}, number={3}, journal={Indonesian Journal of Electrical Engineeri... |
ser | banglaser | BanglaSER | 1,467 recordings | 1,467 | CC BY 4.0 | 2,022 | BanglaSER (Data in Brief 2022) | https://data.mendeley.com/datasets/t9h6p943xy/5 | 2026-07-26T00:00:00 | Realistic Bangla emotional speech from 34 speakers, balanced by gender, across 5 emotions. | @article{Das_2022, title={BanglaSER: A speech emotion recognition dataset for the Bangla language}, volume={42}, ISSN={2352-3409}, url={http://dx.doi.org/10.1016/j.dib.2022.108091}, DOI={10.1016/j.dib.2022.108091}, journal={Data in Brief}, publisher={Elsevier BV}, author={Das, Rakesh Kumar and Islam, Nahidul and Ahmed,... |
ser | kbes | KBES (KUET Bangla Emotional Speech) | 900 recordings | 900 | CC BY 4.0 | 2,023 | KBES (Data in Brief 2023) | https://data.mendeley.com/datasets/vsn37ps3rx/4 | 2026-07-26T00:00:00 | Realistic Bangla emotional speech from 35 actors, with low/high intensity labels for 4 of 5 emotions. | @article{Billah_2023, title={KBES: A dataset for realistic Bangla speech emotion recognition with intensity level}, volume={51}, ISSN={2352-3409}, url={http://dx.doi.org/10.1016/j.dib.2023.109741}, DOI={10.1016/j.dib.2023.109741}, journal={Data in Brief}, publisher={Elsevier BV}, author={Billah, Md. Masum and Sarker, M... |
speech | openslr53 | OpenSLR SLR53 (Large Bengali ASR) | 196k utterances / 229 hrs | 229 | CC BY-SA 4.0 | 2,018 | Google crowdsourced speech | https://openslr.org/53/ | 2026-06-28T00:00:00 | Crowdsourced Bangladeshi Bangla read speech. | @inproceedings{Kjartansson_2018, series={sltu_2018}, title={Crowd-Sourced Speech Corpora for Javanese, Sundanese, Sinhala, Nepali, and Bangladeshi Bengali}, url={http://dx.doi.org/10.21437/SLTU.2018-11}, DOI={10.21437/sltu.2018-11}, booktitle={6th Workshop on Spoken Language Technologies for Under-Resourced Languages (... |
speech | commonvoice | Common Voice (bn) | 1,200+ hrs recorded | 1,200 | CC0 | 2,024 | Mozilla Common Voice | https://commonvoice.mozilla.org/bn | 2026-07-01T00:00:00 | Continuously growing crowdsourced speech corpus. | @inproceedings{ardila-etal-2020-common,
title = "Common Voice: A Massively-Multilingual Speech Corpus",
author = "Ardila, Rosana and
Branson, Megan and
Davis, Kelly and
Kohler, Michael and
Meyer, Josh and
Henretty, Michael and
Morais, Reuben and
Saunders, Lindsay... |
speech | oodspeech | OOD-Speech | 1,177 hrs | 1,177 | CC BY-NC 4.0 | 2,023 | OOD-Speech (2023) | https://bengaliai.github.io/asr | 2026-07-19T00:00:00 | Large-scale out-of-distribution Bangla ASR benchmark. | @misc{rakib2023oodspeechlargebengalispeech,
title={OOD-Speech: A Large Bengali Speech Recognition Dataset for Out-of-Distribution Benchmarking},
author={Fazle Rabbi Rakib and Souhardya Saha Dip and Samiul Alam and Nazia Tasnim and Md. Istiak Hossain Shihab and Md. Nazmuddoha Ansary and Syed Mobassir Hossen ... |
speech | lipighor | Lipi-Ghor (bn-882-SSTT) | 882 hrs | 882 | CC BY 4.0 | 2,026 | Lipi-Ghor-882 (DL Sprint 4.0, Team Villagers) | https://huggingface.co/datasets/Sanjidh090/Lipi-Ghor-bn-882-SSTT | 2026-07-22T00:00:00 | Long-form multi-speaker Bangla ASR and speaker-diarization corpus with aligned transcriptions and timestamps. Gated access. | @misc{hasan2026makehardheareasy,
title={Make It Hard to Hear, Easy to Learn: Long-Form Bengali ASR and Speaker Diarization via Extreme Augmentation and Perfect Alignment},
author={Sanjid Hasan and Risalat Labib and A H M Fuad and Bayazid Hasan},
year={2026},
eprint={2602.23070},
archivePre... |
speech | bengali-loop-asr | Bengali-Loop (ASR corpus) | 191 recordings / 158.6 hrs | 158.6 | Mixed | 2,026 | Bengali-Loop (DL Sprint 4.0, Tabib et al.) | https://www.kaggle.com/competitions/dl-sprint-4-0-bengali-long-form-speech-recognition/data | 2026-08-10T00:00:00 | Long-form Bangla ASR corpus of 792k words from 11 YouTube channels, built by subtitle extraction with human-in-the-loop transcript verification. | @misc{tabib2026bengaliloopcommunitybenchmarkslongform,
title={Bengali-Loop: Community Benchmarks for Long-Form Bangla ASR and Speaker Diarization},
author={H. M. Shadman Tabib and Istiak Ahmmed Rifti and Abdullah Muhammed Amimul Ehsan and Somik Dasgupta and Md Zim Mim Siddiqee Sowdha and Abrar Jahin Sarker ... |
summ | xlsum | XL-Sum (bn) | 10,126 articles | 10,126 | CC BY-NC-SA 4.0 | 2,021 | XL-Sum (Findings ACL 2021) | https://huggingface.co/datasets/csebuetnlp/xlsum | 2026-06-28T00:00:00 | Professional article–summary pairs from BBC Bangla. | @inproceedings{hasan-etal-2021-xl,
title = {{XL}-Sum: Large-Scale Multilingual Abstractive Summarization for 44 Languages},
author = {Hasan, Tahmid and Bhattacharjee, Abhik and others},
booktitle = {Findings of ACL},
year = {2021}
} |
summ | bansdata | BANSData | 19,096 pairs | 19,096 | Research only | 2,021 | Bangla abstractive news summarization | https://www.kaggle.com/datasets/prithwirajsust/bengali-news-summarization-dataset | 2026-02-20T00:00:00 | News summarization pairs crawled from Bangla dailies. | @InProceedings{10.1007/978-981-33-4673-4_4,
author="Bhattacharjee, Prithwiraj
and Mallick, Avi
and Saiful Islam, Md.
and Marium-E-Jannat",
editor="Kaiser, M. Shamim
and Bandyopadhyay, Anirban
and Mahmud, Mufti
and Ray, Kanad",
title="Bengali Abstractive News Summarization (BANS): A Neural Attention Approach",
booktitle... |
summ | banglachqsumm | BanglaCHQ-Summ | 2,350 pairs | 2,350 | CC BY-NC-SA 4.0 | 2,023 | BanglaCHQ-Summ (BLP 2023) | https://github.com/alvi-khan/BanglaCHQ-Summ | 2026-07-19T00:00:00 | Abstractive summarization dataset for Bangla consumer health questions. | @inproceedings{khan-etal-2023-banglachq,
title = "{B}angla{CHQ}-Summ: An Abstractive Summarization Dataset for Medical Queries in {B}angla Conversational Speech",
author = "Khan, Alvi and
Kamal, Fida and
Chowdhury, Mohammad Abrar and
Ahmed, Tasnim and
Laskar, Md Tahmid Rahman and
... |
textcls | banfakenews | BanFakeNews | ~50k articles | 50,000 | CC BY-NC 4.0 | 2,020 | BanFakeNews (LREC 2020) | https://github.com/Rowan1224/FakeNews | 2026-06-11T00:00:00 | Annotated fake vs. authentic Bangla news articles. | @inproceedings{hossain-etal-2020-banfakenews,
title = "{B}an{F}ake{N}ews: A Dataset for Detecting Fake News in {B}angla",
author = "Hossain, Md Zobaer and
Rahman, Md Ashraful and
Islam, Md Saiful and
Kar, Sudipta",
editor = "Calzolari, Nicoletta and
B{\'e}chet, Fr{\'e}d{\'e}ric ... |
textcls | potrika | Potrika | 665k articles | 665,000 | CC BY 4.0 | 2,022 | Potrika (2022) | https://doi.org/10.17632/v362rp78dc.2 | 2026-01-30T00:00:00 | Large single-label news corpus, 8 categories, 6 dailies. | @misc{ahmad2022potrikarawbalancednewspaper,
title={Potrika: Raw and Balanced Newspaper Datasets in the Bangla Language with Eight Topics and Five Attributes},
author={Istiak Ahmad and Fahad AlQurashi and Rashid Mehmood},
year={2022},
eprint={2210.09389},
archivePrefix={arXiv},
primar... |
textcls | bard | BARD | 376,226 articles | 376,226 | Research only | 2,018 | Bangla Article Dataset | https://github.com/tanvirfahim15/BARD-Bangla-Article-Classifier | 2026-02-14T00:00:00 | News articles across 5 categories. | @inproceedings{Tanvir_Alam_2018, title={BARD: Bangla Article Classification Using a New Comprehensive Dataset}, url={http://dx.doi.org/10.1109/ICBSLP.2018.8554382}, DOI={10.1109/icbslp.2018.8554382}, booktitle={2018 International Conference on Bangla Speech and Language Processing (ICBSLP)}, publisher={IEEE}, author={T... |
textcls | bttc | BTTC | 10,283 messages | 10,283 | CC BY 4.0 | 2,026 | Mendeley Data (2026) | https://doi.org/10.17632/5wrm959d6f.1 | 2026-07-24T00:00:00 | Tri-class Bangla SMS/Telegram corpus for spam, ham, and promotional message classification. | @misc{https://doi.org/10.17632/5wrm959d6f.1,
doi = {10.17632/5WRM959D6F.1},
url = {https://data.mendeley.com/datasets/5wrm959d6f/1},
author = {Tahmid, Labib and Abedin, Istiak and Ahmed, Kawsar and Hassan Chowdhury , Zinedin and Sultana, Babe },
keywords = {Computer Science, Artificial Intelligence, Cybersecuri... |
textcls | banfakenews2 | BanFakeNews-2.0 | 60,000 articles | 60,000 | Apache 2.0 | 2,025 | BanFakeNews-2.0 (IndoNLP 2025) | https://www.kaggle.com/datasets/hrithikmajumdar/bangla-fake-news | 2026-07-26T00:00:00 | Fake-news corpus with 47,000 authentic and 13,000 manually annotated fake articles across 13 categories. | @inproceedings{shibu-etal-2025-scarcity,
title = "From Scarcity to Capability: Empowering Fake News Detection in Low-Resource Languages with {LLM}s",
author = "Shibu, Hrithik Majumdar and
Datta, Shrestha and
Miah, Md. Sumon and
Sami, Nasrullah and
Chowdhury, Mahruba Sharmin and
... |
Bangla NLP Catalog
A machine-readable catalog of Bangla (Bengali) NLP resources: 813 papers, 63 datasets, 20 models, and 9 tools across 26 tasks, each tagged by task and carrying a source link.
This is the data behind BanglaNLP Hub. It is metadata about resources, not the resources themselves: no corpora or model weights are redistributed here, only structured records pointing at them.
Why this exists
Bangla is spoken by roughly 240 million people and is still treated as low-resource in practice, largely because the work that does exist is scattered across venues, personal pages, and dead links. This catalog is an attempt to make it countable and searchable, so that a researcher can see what already exists before rebuilding it.
Configs
| Config | Rows | What a row is |
|---|---|---|
papers |
813 | A published paper, with title, authors, venue, year, link, and task |
datasets |
63 | A dataset, with name, desc, size, license, link, and bibtex where known |
models |
20 | A released model, with arch, params, tasks, stage, and link |
tools |
9 | A library or tool, with lang, install, author, and a verified flag |
from datasets import load_dataset
papers = load_dataset("kishormorol/bangla-nlp-catalog", "papers", split="train")
datasets_ = load_dataset("kishormorol/bangla-nlp-catalog", "datasets", split="train")
How it is maintained
Entries are added and corrected through pull requests on the GitHub repo, where links are checked in CI. This dataset is a snapshot of that catalog; the repo is the source of truth.
Limitations
Coverage is best for tasks with active published work (classification, MT, NER, QA) and thinner elsewhere. license fields reflect what the source stated at the time of entry and should be re-checked before you rely on them. Corrections are welcome as PRs or as a discussion here.
Citation
@software{morol_banglanlp_hub,
author = {Morol, Md Kishor},
title = {BanglaNLP Hub: A Catalog of Bangla NLP Resources},
url = {https://github.com/kishormorol/BanglaNLP-Hub},
license = {MIT}
}
- Downloads last month
- 72