Datasets:

Modalities:
Text
Formats:
json
Languages:
French
Size:
< 1K
ArXiv:
Libraries:
Datasets
pandas
imenelydiaker's picture
Add citation (#7)
7503cb7 verified
metadata
dataset_info: null
splits:
  - name: test
task_categories:
  - question-answering
language:
  - fr
pretty_name: Syntec dataset for information retrieval
configs:
  - config_name: documents
    data_files:
      - split: test
        path: documents.json
  - config_name: queries
    data_files:
      - split: test
        path: queries.json

Syntec dataset for information retrieval

This dataset has been built from the Syntec Collective bargaining agreement. Its purpose is information retrieval.

Dataset Details

The dataset is rather small. It is intended to be used only as a test set, for fast evaluation of models.

It is split into 2 subsets :

  • queries : it features 100 manually created questions. Each question is mapped to the article that contains the answer.
  • documents : corresponds to the 90 articles from the collective bargaining

Usage

import datasets

# Download the documents (corpus)
corpus_raw = datasets.load_dataset("lyon-nlp/mteb-fr-retrieval-syntec-s2p", "documents")

# Download the queries
queries_raw = datasets.load_dataset("lyon-nlp/mteb-fr-retrieval-syntec-s2p", "queries")

Dataset Description

The collective bargaining agreement is applicable to employees of Technical Design Offices, Consulting Engineering Firms and Consulting Companies.

The dataset contains 100 questions, each having their answer in 1 of the 90 articles of the documents. The dataset was manually annotated. It's small size allows for quick prototyping.

  • Curated by: Wikit AI (https://www.wikit.ai/)
  • Language(s) (NLP): French
  • License: [More Information Needed]

Dataset Sources

https://www.syntec.fr/

Citation

If you use this dataset in your work, please consider citing:

@misc{ciancone2024extending,
      title={Extending the Massive Text Embedding Benchmark to French}, 
      author={Mathieu Ciancone and Imene Kerboua and Marion Schaeffer and Wissam Siblini},
      year={2024},
      eprint={2405.20468},
      archivePrefix={arXiv},
      primaryClass={cs.CL}
}

Contact

mathieu@wikit.ai marion@wikit.ai