[Esip-discovery] ESIP Discovery Cluster: Follow up from today's meeting

[Sara Lafia]

Hello Discovery Cluster,

Thanks for a great discussion! As promised, I'm sending along a few science
of science articles to spark our collective curiosity:

   - Measuring the Evolution of a Scientific Field through Citation Frames
   <https://www-cs.stanford.edu/~srijan/pubs/paper-citations.pdf>
   - Discovering discoveries: Identifying biomedical discoveries using
   citation contexts
   <https://www.sciencedirect.com/science/article/abs/pii/S1751157716302255>
   - Design and Update of a Classification System: The UCSD Map of Science
   <https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0039464>

I also wanted to add that multi-level topic modeling to cluster similar
publications might be worthwhile to explore down the line for the UBD tool.
Here's an example of a research map I made with this technique; the
position of each document is based on its multiple topics, which are
emergent and are labeled by the people who authored the articles (
https://eri-research-dashboard.herokuapp.com/)

Chris and I also worked with Megan Carter to publicize efforts to
contribute to the Usage-Based Discovery tool (
https://www.esipfed.org/collaboration-updates/linking-datasets-to-applications
).

Finally, I participated in the Coleridge Kaggle competition mentioned
earlier (https://www.kaggle.com/c/coleridgeinitiative-show-us-the-data/)
and had some luck training a transformer-based NER model with spaCy
<https://spacy.io/usage/training> to detect dataset references. My research
is ongoing and I'll keep the group posted on progress and useful takeaways.

Best wishes,


-- 
*Sara Lafia, Ph.D.*
Research Fellow, ICPSR
saralafia.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.esipfed.org/pipermail/esip-discovery/attachments/20210916/719a9a01/attachment.htm>