Skip to main content
Responsible Operations: Data Science, Machine Learning and AI in Libraries

Responsible Operations: Data Science, Machine Learning and AI in Libraries

February 2020

Responsible Operations: Data Science, Machine Learning, and Artificial Intelligence in Libraries
Thomas Padilla, Practitioner Researcher in Residence, OCLC

Implementation and use of data science, machine learning and artificial intelligence (AI) in the context of the library is still very much in its early stages. This 2019 report, commissioned by OCLC and authored by Thomas Padilla as Practitioner Researcher in Residence, represents an agenda for moving forward. What makes the report noteworthy is the recognition by its author that successfully doing so requires an integrated approach. In meeting the challenges of new tasks and technologies, the approach should be holistic in establishing new library policies, roles, and workflows.

Padilla’s report looks at seven high-level categories with challenges and recommendations noted for each. Based on input from an impressive advisory group as well as interviews and face-to-face engagement with a “landscape” group, the seven areas are:

  • Committing to Responsible Operations
  • Description and Discovery
  • Shared Methods and Data
  • Machine-Actionable Collections
  • Workforce Development
  • Data Science Services
  • Sustaining Interprofessional and Interdisciplinary Collaboration

The first bullet above refers to a concept articulated by Rumman Chowdhury, presented in the report as follows: “... responsible operations refers to individual, organizational, and community capacities to support responsible use of data science, machine learning, and AI.” There is an assumption of shared ethical commitment on the part of those individuals, organizations, and communities. Ethical efforts recognize a need for managing bias as well as transparency, explainability, and accountability. Responsible operations rely on data science fluency and on the availability of tools “designed and documented in such a way that they make it possible for users of varying skill levels to contribute.”

Sections on the subsequent bullet items address applications of technologies (data science, machine learning, and AI) that would enhance the effectiveness and efficiencies of library services and on skills and practices that would be needed in the workforce. Recommendations for next steps in each of those areas included formation of working groups, developing collaborative partnerships, and cross-functional community studies although no timeframe was specified for next steps. The report’s closing paragraph notes, “With investigations that span collections and research support, the scholarly record, the system-wide library, user studies, community catalysts, and data science, OCLC Research will assess how it might best contribute. This assessment will take place in light of our capacity, existing research library community efforts and expressed community needs.”

The full text of the report (PDF file format) may be downloaded here from the OCLC web site. NISO’s Director of Content, Jill O’Neill conversed with Thomas Padilla about the report in an interview here.