datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
maltese-ocr-corpus
Real Maltese OCR Paragraph Pairs
Paragraph-level (image, text) pairs from real, digitally-typeset
Maltese PDFs. Built to close the synthetic-only gap in Maltese OCR
training data - see the accompanying paper (LV-ROVER-MLT, DocEng 2026)
for context: the paper and the corpus-building scripts
(package_for_hf.py, align_pdf_paragraphs.py, under
experiments/neural_resume/corpus/) are at
github.com/adamd1985/doceng2026.
The frozen competition submission (the Tesseract LV-ROVER-MLT… See the full description on the dataset page: https://huggingface.co/datasets/radmada/maltese-ocr-corpus.maltese_news_categories
Maltese News Categories
A multi-label topic classification dataset for Maltese News Articles.
Data Collection
The data was collected from the press_mt subset from Korpus Malti v4.0.
Article contents were cleaned to filter out JavaScript, CSS, & repeated non-Maltese sub-headings.
The labels are based on the category field from this corpus.
Additional filtering & cleaning was performed as follows:
Documents with generic categories (News, Local, Headlines, Uncategorised… See the full description on the dataset page: https://huggingface.co/datasets/MLRS/maltese_news_categories.MalteseNewsClassification
MalteseNewsClassification
An MTEB dataset
Massive Text Embedding Benchmark
A multi-label topic classification dataset for Maltese News
Articles. The data was collected from the press_mt subset from Korpus
Malti v4.0. Article contents were cleaned to filter out JavaScript, CSS,
& repeated non-Maltese sub-headings. The labels are based on the category
field from this corpus.
Task category
t2c
Domains
Constructed, Written
Reference… See the full description on the dataset page: https://huggingface.co/datasets/mteb/MalteseNewsClassification.maltese_embeddings
Dataset Summary
This repository contains three distinct datasets focusing on Maltese word embeddings:
GloVe Maltese Word Embeddings
Embeddings generated using GloVe on the "korpus_malti" dataset, the largest Maltese corpus available.
Word2Vec Maltese Word Embeddings
Word embeddings for Maltese obtained using Word2Vec trained on the "korpus_malti" dataset.
PPMI Maltese Word Embeddings
Pointwise Mutual Information (PPMI) based word embeddings generated from ConceptNet data via SVD… See the full description on the dataset page: https://huggingface.co/datasets/DGurgurov/maltese_embeddings.maltese_news_headlines
Maltese News Headlines
A headline-article pairs dataset for Maltese News Articles.
This dataset is intended to be used for headline generation from the article content.
Data Collection
The data was collected from the press_mt subset from Korpus Malti v4.0.
Article contents were cleaned to filter out JavaScript, CSS, & repeated non-Maltese sub-headings.
The title and base URL features are based on the title & url fields from this corpus, respectively.… See the full description on the dataset page: https://huggingface.co/datasets/MLRS/maltese_news_headlines.mapa_maltese
MAPA Maltese
Named-Entity Recognition dataset from the MAPA Project.
This dataset has some fixes as detailed in Cross-Lingual Transfer from Related Languages: Treating Low-Resource Maltese as Multilingual Code-Switching:
Manually fixed some inconsistencies between Level 1 & Level 2 tags.
Manually added the labels for some spans which were marked as entity spans but didn't have the tags.
Manually fixed incorrectly marked spans with respect to tokenisation (either having a sub-word… See the full description on the dataset page: https://huggingface.co/datasets/MLRS/mapa_maltese.maltese-speech-datasetMaltese_sentimentmaltese_sa
Sentiment Analysis Data for the Maltese Language
Dataset Description:
This dataset contains sentiment analysis data originating from comments on news articles and social media posts. It combines two datasets from Cortis and Davis (2019) and Dingli and Sant (2016).
Data Structure:
The data was utilized for the project on injecting external commonsense knowledge into multilingual Large Language Models.
Citation:
@inproceedings{cortis-davis-2019-social,
title = "A Social Opinion… See the full description on the dataset page: https://huggingface.co/datasets/DGurgurov/maltese_sa.alpaca_maltese_tacoThis repository contains the dataset used for the TaCo paper.
The dataset follows the style outlined in the TaCo paper, as follows:
{
"instruction": "instruction in xx",
"input": "input in xx",
"output": "Instruction in English: instruction in en ,
Response in English: response in en ,
Response in xx: response in xx "
}
Please refer to the paper for more details: OpenReview
If you have used our dataset, please cite it as follows:
Citation… See the full description on the dataset page: https://huggingface.co/datasets/saillab/alpaca_maltese_taco.maltese_gabra
Dataset Summary
This repository comprises the dataset with the Maltese words extracted from Gabra:
Maltese Words and Their English Glosses from Gabra
Dataset containing Maltese words and their corresponding English glosses, extracted from the Gabra database.
Languages
Maltese
Dataset Creation
The data was extracted from Gabra for further use in training PPMI embeddings.
Contributors
Daniil Gurgurov
maltese_conceptnet
ConceptNet Data for the Maltese Language
Dataset Description:
This dataset contains data extracted from ConceptNet using the dedicated module for fetching knowledge from the graph, available on GitHub.
Data Structure:
The data is converted from triplets into natural text using a pre-defined relationship mapping and split into training and validation sets. It was used for training language adapters for the project aimed at injecting external commonsense knowledge into multilingual… See the full description on the dataset page: https://huggingface.co/datasets/DGurgurov/maltese_conceptnet.PAWS-X-maltese
Dataset Card for "PAWS-X-maltese"
More Information needed
maltese-news-nli-random
Dataset Card for "maltese-news-nli-random"
More Information needed
maltese-news-nli-sports
Dataset Card for "maltese-news-nli-sports"
More Information needed
alpaca-maltese-cleanedThis repository contains the dataset used for the TaCo paper.
Please refer to the paper for more details: OpenReview
If you have used our dataset, please cite it as follows:
Citation
@inproceedings{upadhayay2024taco,
title={TaCo: Enhancing Cross-Lingual Transfer for Low-Resource Languages in {LLM}s through Translation-Assisted Chain-of-Thought Processes},
author={Bibek Upadhayay and Vahid Behzadan},
booktitle={5th Workshop on practical ML for limited/low resource settings, ICLR},
year={2024}… See the full description on the dataset page: https://huggingface.co/datasets/saillab/alpaca-maltese-cleaned.