ub-mannheim/nfdi

Basic info about NFDI consortia. Named entity linking for them. Support for Wikidata WikiProject NFDI.

NFDI

The Python library NFDI provides:

• basic information about NFDI (also known as Nationale ForschungsDatenInfrastruktur and National Research Data Infrastructure) and all funded NFDI consortia,
• support for Wikidata WikiProject NFDI aimed to create and edit the Wikidata entities and entity schemas relevant for NFDI,
• simple named entity linker on texts for the accepted NFDI consortia.

The NFDI Jupyter Book describes:

• how to use the library,
• how to send SPARQL queries to Wikidata and to get visualisations for NFDI consortia,
• Wikidata WikiProject NFDI and relevant entity schemas,
• how we parsed the data,
• how we edited Wikidata.

Table of contents

• Installation
• How to use
• NFDI Jupyter Book

Installation

pip install NFDI

How to use

Module info

[docs module info]

The module info has classes consortium and consortia:

from nfdi import info
nfdi = info.consortia()
berd = info.consortium('BERD@NFDI')

The instance nfdi has the following attributes: 'label', 'homepage', 'wikidata', 'github', 'google', 'linkedin', 'twitter', 'youtube', 'zenodo', 'labels', 'consortia'. For example:

nfdi.twitter

prints

"https://twitter.com/NFDI_de"

The instances nfdi and berd have the methods print, dict and _wikidata:

json = berd._wikidata()
nfdi.print()

The json variable contains JSON representation of the corresponding entity at Wikidata including labels, aliases and descriptions:

print('LABELS', json.get('labels'))
print('DESCRIPTIONS', json.get('descriptions'))
print('ALIASES', json.get('aliases'))

It prints:

LABELS {'en': {'language': 'en', 'value': 'BERD@NFDI'}, 'de': {'language': 'de', 'value': 'BERD@NFDI'}, 'fr': {'language': 'fr', 'value': 'BERD@NFDI'}, 'bar': {'language': 'bar', 'value': 'BERD@NFDI'}, 'de-at': {'language': 'de-at', 'value': 'BERD@NFDI'}, 'de-ch': {'language': 'de-ch', 'value': 'BERD@NFDI'}, 'de-formal': {'language': 'de-formal', 'value': 'BERD@NFDI'}, 'en-ca': {'language': 'en-ca', 'value': 'BERD@NFDI'}, 'en-gb': {'language': 'en-gb', 'value': 'BERD@NFDI'}, 'es': {'language': 'es', 'value': 'BERD@NFDI'}, 'nl': {'language': 'nl', 'value': 'BERD@NFDI'}, 'pt': {'language': 'pt', 'value': 'BERD@NFDI'}, 'simple': {'language': 'simple', 'value': 'BERD@NFDI'}}
DESCRIPTIONS {'en': {'language': 'en', 'value': 'NFDI consortium for Business, Economic and Related Data (Social and Behavioural Sciences)'}, 'de': {'language': 'de', 'value': 'NFDI für Wirtschaftsdaten und Verwandtes (Sozial- und Verhaltenswissenschaften)'}}
ALIASES {'en': [{'language': 'en', 'value': 'BERD-NFDI'}], 'de': [{'language': 'de', 'value': 'BERD-NFDI'}], 'fr': [{'language': 'fr', 'value': 'BERD-NFDI'}], 'bar': [{'language': 'bar', 'value': 'BERD-NFDI'}], 'de-at': [{'language': 'de-at', 'value': 'BERD-NFDI'}], 'de-ch': [{'language': 'de-ch', 'value': 'BERD-NFDI'}], 'de-formal': [{'language': 'de-formal', 'value': 'BERD-NFDI'}], 'en-ca': [{'language': 'en-ca', 'value': 'BERD-NFDI'}], 'en-gb': [{'language': 'en-gb', 'value': 'BERD-NFDI'}], 'es': [{'language': 'es', 'value': 'BERD-NFDI'}], 'nl': [{'language': 'nl', 'value': 'BERD-NFDI'}], 'pt': [{'language': 'pt', 'value': 'BERD-NFDI'}], 'simple': [{'language': 'simple', 'value': 'BERD-NFDI'}]}

Module data

[docs module data]

The module data has raw data as a dictionary:

from nfdi import data
data.raw()

Module nel

[docs module nel]

The module nel provides simple rule-based named entity linker for the NFDI consortia. In Jupyter Notebook use

from nfdi.nel import linker, test
t = linker(test)
t.render()

where test stores the following sentences:

What are BERD@NFDI, NFDI4Earth, NFDI4DataScience, NFDI-MatWerk, PUNCH4NFDI, FAIRmat and Text+?
How are they related to NFDI4Ing, NFDI4Culture, NFDI4Chem and NFDIGHGA?

In Python console use:

from nfdi.nel import linker, test
t = linker(test)
t.serve()

The Wikidata QIDs are stored in .ent_id_:

from nfdi.nel import linker, test
t = linker(test)
for span in t.doc.ents:
    print((span.text, span.ent_id_, span.label_))

It prints:

('BERD@NFDI', 'Q108542181', 'ORG')
('NFDI4Earth', 'Q108542504', 'ORG')
('NFDI4DataScience', 'Q108542422', 'ORG')
('NFDI-MatWerk', 'Q108542607', 'ORG')
('PUNCH4NFDI', 'Q108542637', 'ORG')
('FAIRmat', 'Q108542373', 'ORG')
('Text+', 'Q98271443', 'ORG')
('NFDI4Ing', 'Q98380344', 'ORG')
('NFDI4Culture', 'Q98276929', 'ORG')
('NFDI4Chem', 'Q96678459', 'ORG')
('NFDIGHGA', 'Q98380337', 'ORG')

NFDI Jupyter Book

Check out NFDI Jupyter Book.

Deploying the Book locally

First, create and activate conda environment using the provided docs/environment.yml:

conda env create -f docs/environment.yml
conda activate nfdi

Then, build the Book:

jb build docs

Open the file docs/_build/html/index.html in a browser.

To remove the build folder, run:

jb clean --all docs

Deploying the Book at GitHub

Install ghp-import:

pip install ghp-import

Once the book is built, run:

ghp-import -n -p -f docs/_build/html