azugarini/crossword-clues-QA

ECWCA-Educational CrossWord Clues Answering A CALAMITA Challenge

Overview

ECWCA (Educational CrossWord Clues Answering) is a dataset designed to evaluate the knowledge and reasoning capabilities of Large Language Models (LLMs) through crossword clue-answering tasks. The dataset consists of clues and their respective answers, constructed based on entities and facts extracted from Italian Wikipedia. ECWCA is part of the CALAMITA Challenge.

Dataset Description

Data Format: The dataset consists of paired crossword clues (questions) and their expected answers.

Language: The dataset focuses on Italian entities and facts, making it particularly relevant for research in Italian NLP.

Usage

The dataset can be accessed and loaded via the datasets library in Python:

from datasets import load_dataset
dataset = load_dataset("azugarini/crossword-clues-QA")

This dataset is particularly useful for NLP researchers working on challenging linguistic tasks requiring deep understanding of word meanings and relationships.

Citation

If you use this dataset in your research, please cite both:

@inproceedings{zugarini2024ecwca,
  title={ECWCA-Educational CrossWord Clues Answering A CALAMITA Challenge},
  author={Zugarini, Andrea and Zeinalipour, Kamyar and Fusco, Achille and Zanollo, Asya},
  booktitle={Proceedings of the 10th Italian Conference on Computational Linguistics (CLiC-it 2024), Pisa, Italy},
  year={2024}
}

@inproceedings{attanasio2024calamita,
  title={CALAMITA: Challenge the Abilities of LAnguage Models in ITAlian},
  author={Attanasio, Giuseppe and Basile, Pierpaolo and Borazio, Federico and Croce, Danilo and Francis, Maria and Gili, Jacopo and Musacchio, Elio and Nissim, Malvina and Patti, Viviana and Rinaldi, Matteo and others},
  booktitle={Proceedings of the 10th Italian Conference on Computational Linguistics (CLiC-it 2024), Pisa, Italy},
  year={2024}
}