Curation and structuring of the scientific production of the IATEXT in Wikidata
Wikidata, WikiCite and Scholia as tools to elaborate a corpus of linked bibliographic data
Iván Hernández-Cazorla
Manuel Ramírez-Sánchez
Gregorio Rodríguez-Herrera
A bit of history
Wikidata-IATEXT
Motivation
Open Science and Open Data
Corpus of linked data
↓
Scientific profiles
↔
Bibliographical data
Previous state
| Around... | |
| 4300 research institutes in Wikidata. | 86 research institutes with at leat 1 researcher linked with the property «member of» (P463). |
| 25 research institutes with at least 1 publication linked to one of its members. | |
Current state
Methodology
| Phases | |
| Researchers | Publications |
| Useful and necessary items to link between them and enrich the data | |
| Main sources used as references | ||
| Websites | Dialnet | Acceda |
| Mostly lists without a defined structure | Standarized model. Can't make an automatic extraction |
Potential repository. Weak structure and data |
Results
Useful tools to get the results:
- Listeria and the Wikidata Query Service.
exporter.py, developed script to extract the data in RDF/Turtle. Usewikidata-clito perform the queries.
In addition to the previously:
- 43 429 editions.
- 1689 items edited, from which 1430 created.
- From the previous number, 1080 are publications by members of the IATEXT.
Conclusions
Scholia: data visualization of the items inserted to the bibliographical corpus
All the researchers can be visualized.
But there isn't time available to add all the publications.
Pretty complete panels for the organization (Q27639076) and authors.
Weaker citations panel for both:
Only 7 articles with citations
About citations
We didn't plan to add citations during this project and we didn't add references.
Who linked these items together with P2860 ("Reference")?
+
How can this corpus be used?
Presentation available in https://ivanhercaz.com/p/talks.html
Made with reveal.js (JavaScript framework developed by Hakim El Hattab).