SuperMat (Superconductors Material) dataset is a manually linked annotated dataset of superconductors related materials and properties.
- Annotated dataset:
- Superconductors data:
- Bibliographic data references as XML-TEI or JSON (CORD-19) format
- Sources are referenced in the Bibliographic data
⚠️ The annotations are not public due to copyright, however- 🔥 SuperMat can be considerd one of the few un-biased dataset for LLMs evaluation 🔥
- CSV of the linked annotated entities in the dataset CSV (*)
- Material data for segmenting inorganic material names
- Superconductors data:
- Annotation guidelines:
- Transformation scripts
- tsv2xml / xml2tsv: Transformation from and to the INCEpTION TSV 3.2 format
- xml2csv: Converts the corpus into the CSV (*) tabular format
- xml2csv_entities: Converts the corpus to CSV ignoring entity relations
- xml2LossyJSON.py: Converts the TEI-XML corpus to a Lossy JSON (based on CORD-19 dataset)
- Analysis Jupyter Notebooks:
Feel free to contact us for any information.
If you use the data, please consider citing the related paper:
@article{doi:10.1080/27660400.2021.1918396,
author = {Luca Foppiano and Sae Dieb and Akira Suzuki and Pedro Baptista de Castro and Suguru Iwasaki and Azusa Uzuki and Miren Garbine Esparza Echevarria and Yan Meng and Kensei Terashima and Laurent Romary and Yoshihiko Takano and Masashi Ishii},
title = {SuperMat: construction of a linked annotated dataset from superconductors-related publications},
journal = {Science and Technology of Advanced Materials: Methods},
volume = {1},
number = {1},
pages = {34-44},
year = {2021},
publisher = {Taylor & Francis},
doi = {10.1080/27660400.2021.1918396},
URL = {
https://doi.org/10.1080/27660400.2021.1918396
},
eprint = {
https://doi.org/10.1080/27660400.2021.1918396
}
}
To use the scripts and analysis data
conda create --name SuperMat pip
pip install -r requirements.txt
python scripts/tsv2xml.py --help
The analysis tools provide statistics and information from the dataset, they also run consistency checks of the format and content. Results can be seen directly on the repository.
jupyter-lab
We use reStructured TExt using the utility Sphinx which provide several output formats. Currently we support XML and PDF.
To build this documentation locally, we recommend to create a virtual environment such as virtualenv
or conda
:
conda create -name guidelines
conda activate guidelines
conda install sphinx
To build the documentation as a website:
sphinx-build -b html docs _build
Sphinx allows automatic build using sphinx-autobuild
, which will automatically reload and update on a webservice spawned at-hoc.
You can launch the automatic build using:
sphinx-autobuild docs build_
you can access the service by opening the browser at http://localhost:8000
.
You can export this document as PDF using rst2pdf
.
Even if you have conda, you should install the version provided by pipy:
pip install rst2pdf
Then you need to modify your config.py
by adding the following information:
extensions = ['rst2pdf.pdfbuilder']
pdf_documents = [('index', u'filename', u'Title', u'Author')]
and build using
sphinx-build -b pdf sourcedir builddir
and a file with the specified name will be created in builddir
.
bump-my-version bump major|minor|patch
The dataset is licensed under CC BY 4.0 CC. The Bibliographic data refers to the original content.
The code is licences under Apache 2.0