A presentation at the joint Upper Midwest Digital Collections Conference and Minnesota Digital Library Annual Meeting in 2020.
The diversity of a digital collection is often assessed by considering the diversity of its content. In order for collections to be truly inclusive, however, they need to emphasize usability alongside broad representation. The University of Cincinnati Libraries discusses how diversity and accessibility are intersectional considerations of digital collections, and introduces newly implemented workflows and standards designed to create accessible, inclusive digital collections that broaden usability for all.
Presentation recording available (Starts at 14 minutes, 30 seconds): https://youtu.be/srIPaD7RvYo
The files in this work represent the presentations and workshop content from the 5th UC Data Day held 2020-10-23.
The theme was “World Changing Data: How Digital Data Will Change Our Future”.
The Keynote speaker was Glenn Ricart, of US Ignite - "Smart Runs on Data"
Interactive Panel featuring: Michael Dunaway (moderator) - Whitney Gaskins (Asst Dean, CEAS - Incl Excellence & Comm Engagmnt) - Zvi Biener (Assoc Professor, A&S Philosophy) - Prashant Khare (Asst Professor, CEAS - Aerospace Eng & Eng Mechanics)- Sam Anand (Professor, CEAS - Mechanical Eng) - Achala Vagal
(Professor Clinical - GEO, COM Radiology Neuroradiology)
Power Sessions:
George Turner - Indiana University - High-Performance Computing at UC
Erin McCabe - University of Cincinnati - Text Mining, Natural Language Processing & AI
link to slides - https://bit.ly/dataday_slides
link to code - https://bit.ly/dataday_code
Videos of the day can be found on the UC Libraries STRC1 youtube channel - https://www.youtube.com/c/STRC1/videos
The data sets were derived from coronavirus related scientific literature using the CORD-19 dataset released by the Allen Institute of Artificial Intelligence as of July 14, 2020, using the Elasticsearch engine hosted by the Digital Scholarship Center (DSC). Through indexing the full-text and the metadata of the article corpus, the research team generated a full-corpus model and 7 different models corresponding to key viral outbreaks from the past several decades' coronaviruses (SARS-CoV, MERS-CoV, and SARS- CoV-2) and non-coronaviruses (HIV, Zika, H1N1, and Ebola). The targeted subsets of the articles used two or more occurrences of virus-specific keywords drawn from conventions established by the World Health Organization.
The data sets were derived from coronavirus related scientific literature using the CORD-19 dataset released by the Allen Institute of Artificial Intelligence as of July 14, 2020, using the Elasticsearch engine hosted by the Digital Scholarship Center (DSC). Through indexing the full-text and the metadata of the article corpus, the research team generated a full-corpus model and 7 different models corresponding to key viral outbreaks from the past several decades' coronaviruses (SARS-CoV, MERS-CoV, and SARS- CoV-2) and non-coronaviruses (HIV, Zika, H1N1, and Ebola). The targeted subsets of the articles used two or more occurrences of virus-specific keywords drawn from conventions established by the World Health Organization.
The data sets were derived from coronavirus related scientific literature using the CORD-19 dataset released by the Allen Institute of Artificial Intelligence as of July 14, 2020, using the Elasticsearch engine hosted by the Digital Scholarship Center (DSC). Through indexing the full-text and the metadata of the article corpus, the research team generated a full-corpus model and 7 different models corresponding to key viral outbreaks from the past several decades' coronaviruses (SARS-CoV, MERS-CoV, and SARS- CoV-2) and non-coronaviruses (HIV, Zika, H1N1, and Ebola). The targeted subsets of the articles used two or more occurrences of virus-specific keywords drawn from conventions established by the World Health Organization.
