Scientific Libraries and Knowledge Infrastructure in the Age of Artificial Intelligence

This page presents the research project entitled “Digital Resource Management Models in Research Libraries: Repository Architecture, Digitization, Research Output Registration, and the Potential of Large Language Models (LLMs) for Resource Exploration,” carried out in 2026 by researchers from the Institute of Information Sciences at the University of the National Education Commission, Krakow, in cooperation with the Jagiellonian Library and international partners.

Scientific Libraries and Knowledge Infrastructure in the Age of Artificial Intelligence

ABOUT THE PROJECT

The project examines the impact of artificial intelligence on the information infrastructure of research libraries. It covers digital resource management, repositories and digital libraries, digitization, CRIS systems, metadata standards, and the long-term preservation and accessibility of resources.

Particular attention is paid to assessing applications of large language models (LLMs) in semantic search, full-text exploration, metadata analysis and quality control, and natural-language access to scholarly resources.

Knowledge infrastructure is studied from the perspectives of users, creators of descriptions and data, and curators responsible for data quality and long-term preservation. The Jagiellonian Library is the project’s main Polish research centre, while its international component was conducted at the University of Amsterdam and selected institutions in Amsterdam.

FUNDING

The Amsterdam component was funded by the University of the National Education Commission, Krakow, through the IDUB-03 “Excellent Mobility” programme and the Vice-Rector for Research Fund. The internship at the Jagiellonian Library is conducted at no cost to the Library.

RESEARCH OBJECTIVES AND AREAS

  1. To examine the architecture of institutional repositories, digital libraries, and resource storage systems.
  2. To analyse digitization processes and standards for describing, processing, and publishing digital materials.
  3. To analyse CRIS systems and their integration with repositories and external scholarly information sources, particularly ORCID, OpenAlex, and OpenAIRE.
  4. To assess mechanisms ensuring the durability, security, and continuity of access to digital resources.
  5. To evaluate the potential of LLMs for semantic search, metadata analysis, full-text exploration, and data quality control.
  6. To identify the organizational, legal, and ethical conditions for the responsible use of artificial intelligence in research libraries.
  7. To design and conduct an application experiment using data from the Jagiellonian University Repository and CRIS system.

PROJECT ORGANIZATION AND STAGES

The project comprises four interrelated stages:

  1. A research internship at the Jagiellonian Library, conducted from 14 April to 15 November 2026, with a break in July and August. Dr Leszek Szafrański serves as the academic supervisor. The internship is conducted through research visits at no cost to the Jagiellonian Library.
  2. An international research component in Amsterdam, conducted from 29 June to 12 July 2026. It included a stay at the University of Amsterdam, participation in the Digital Methods Summer School 2026, study visits, expert consultations, and the collection of comparative material.
  3. An application experiment using LLMs, planned for the second stage of the internship at the Jagiellonian Library.
  4. Synthesis of the findings and preparation of a research report and joint scholarly publication.

PRINCIPAL RESEARCHERS

Principal Researcher 1

Władysław Marek Kolasa, PhD, DSc, Associate Professor at the University of the National Education Commission, Krakow, is a researcher at the Institute of Information Sciences. His contribution includes project coordination, analysis of scholarly information infrastructure, institutional repositories and CRIS systems, research into LLM applications, and preparation of the application experiment. From 29 June to 12 July 2026, he conducted the University of Amsterdam component as an Affiliated Researcher at the Amsterdam School for Cultural Analysis (ASCA).

Principal Researcher 2

Michał Rogoż, PhD, DSc, Associate Professor at the University of the National Education Commission, Krakow, is a researcher at the Institute of Information Sciences. His contribution includes the user perspective on knowledge infrastructure, content analysis, resource discovery, access and use, and the evaluation of information functionality. From 5 to 9 July 2026, he participated in study visits and expert consultations in Amsterdam.

Academic supervisor of the internship at the Jagiellonian Library

Dr Leszek Szafrański – Jagiellonian Library.

AMSTERDAM COMPONENT

From 29 June to 12 July 2026, Władysław Marek Kolasa conducted the international component of the internship at the University of Amsterdam as an Affiliated Researcher at the Amsterdam School for Cultural Analysis (ASCA), hosted by Prof. Richard Rogers. The stay included participation in the Digital Methods Summer School 2026 and research on repositories, CRIS systems, and scholarly data infrastructure.

Michał Rogoż participated in study visits and consultations in Amsterdam from 5 to 9 July 2026, focusing on the user perspective and access to collections. The Digital Methods Summer School formed the methodological part of Kolasa’s stay; the project completed during the summer school is described on the separate “Media and Journalism in the Age of Artificial Intelligence” project page.

CONSULTATIONS AND COOPERATION

The project was prepared and conducted with support from consultations with Dr Leszek Szafrański at the Jagiellonian Library on 16 March, 7 May, and 16 June 2026. The discussions concerned the project assumptions, organization of the internship, progress of the research, and preparation of its Amsterdam component.

On 7 July 2026, a consultation was held with Prof. Paul Groth of the Data Science Centre / Intelligent Data Engineering Lab at the University of Amsterdam. The discussion concerned knowledge graphs, the use of LLMs in data engineering, RAG architectures, and the reliability and citability of AI-generated answers.

On 9 July 2026, a consultation was held with Manuel Llano and Hanneke at the Amsterdam University Library. Topics included the Elsevier Pure-based CRIS system, metadata workflows and validation, integration with SAP and ORCID, and the use of OpenAlex.

Working contacts were also established with representatives of the International Institute of Social History, Allard Pierson, the University of Amsterdam Library, Openbare Bibliotheek Amsterdam, UvA AI Chat/ICTS, and other units working with repositories, CRIS systems, metadata, digital collections, knowledge graphs, and AI-based services.

After the stay, correspondence with specialists and representatives of Amsterdam institutions continued. Additional information and materials were obtained on repositories, metadata management, digital curation, and the development of AI-supported library services.

FINDINGS TO DATE

An important outcome of the Amsterdam component was the examination of SURF Open Research Information, an open infrastructure integrating data from sources including OpenAlex, OpenAIRE, Dutch repositories and CRIS systems, and OpenAPC. Its agent-based solutions allow users to query scholarly data in natural language using LLMs and the Model Context Protocol. On 27 July 2026, analysis and testing began with a view to using the system as a model for the experiment planned at the Jagiellonian Library.

Five main areas of LLM application in scholarly information infrastructure were identified:

  1. Natural-language searching in repositories and CRIS systems.
  2. Metadata quality control and enrichment.
  3. Support for describing archival and library materials.
  4. Publication of CRIS metadata as open linked data.
  5. Development of internal AI tools and principles for their responsible use.

Persistent identifiers, standard formats and vocabularies, APIs, linked data, appropriate staffing, and a responsible AI policy were identified as the fundamental conditions for “LLM readiness.”

PLANNED OUTPUTS

  • a description of digital resource management models in research libraries;
  • identification of organizational and technological good practices;
  • an assessment of potential LLM applications in institutional repositories and CRIS systems;
  • an application experiment using data from the Jagiellonian University Repository and CRIS system;
  • a comparison of selected Polish and Dutch solutions;
  • a research report;
  • a joint scholarly article by Władysław Marek Kolasa and Michał Rogoż.

The application experiment, report, and article are planned outputs and had not been completed as of the date of this page’s latest update.

UPDATES

  • 2026-03-16 – A preparatory consultation was held with Dr Leszek Szafrański at the Jagiellonian Library concerning the assumptions, scope, and organization of the planned research internship.
  • 2026-03-26 – Approval was obtained for a joint research internship by Władysław Marek Kolasa and Michał Rogoż at the Jagiellonian Library under the academic supervision of Dr Leszek Szafrański.
  • 2026-04-02 – Władysław Marek Kolasa was admitted to the Digital Methods Summer School 2026 at the University of Amsterdam.
  • 2026-04-13 – Prof. Richard Rogers issued formal confirmation of Władysław Marek Kolasa’s admission to the Digital Methods Summer School 2026, held from 29 June to 3 July 2026.
  • 2026-04-14 – The research internship at the Jagiellonian Library began.
  • 2026-04-30 – The first stage of the literature review on digital resource management, repositories, CRIS systems, digitization, data curation, and AI and LLM applications in research libraries was completed.
  • 2026-05-07 – A working consultation was held with Dr Leszek Szafrański during the first stage of the internship at the Jagiellonian Library.
  • 2026-05-26 – Władysław Marek Kolasa was accepted as an Affiliated Researcher at the Amsterdam School for Cultural Analysis (ASCA), University of Amsterdam, for the period from 29 June to 12 July 2026; Prof. Richard Rogers served as the host.
  • 2026-06-16 – A consultation was held with Dr Leszek Szafrański concerning the progress of the internship and preparation of its Amsterdam component.
  • 2026-06-29 to 2026-07-12 – Władysław Marek Kolasa completed a research stay at the University of Amsterdam as an Affiliated Researcher at the Amsterdam School for Cultural Analysis. The stay formed the international component of the internship conducted at the Jagiellonian Library.
  • 2026-06-29 to 2026-07-03 – Władysław Marek Kolasa participated in the Digital Methods Summer School 2026 at the University of Amsterdam. The summer school formed the methodological part of his research stay; the sprint project is described on the separate “Media and Journalism in the Age of Artificial Intelligence” project page – https://szkola-doktorska.uken.krakow.pl/social-media-research-puk-main/
  • 2026-07-03 – Thirteen enquiries were sent to specialists and units at the University of Amsterdam and to Amsterdam libraries and digital heritage institutions.
  • 2026-07-05 to 2026-07-09 – Michał Rogoż participated in study visits and expert consultations forming the Amsterdam component of the joint internship at the Jagiellonian Library.
  • 2026-07-07 – A consultation was held with Prof. Paul Groth of the Data Science Centre / Intelligent Data Engineering Lab at the University of Amsterdam concerning knowledge graphs, the use of LLMs in data engineering, RAG architectures, and the reliability and citability of AI-generated answers.
  • 2026-07-09 – A consultation was held with Manuel Llano and Hanneke at the Amsterdam University Library concerning the Elsevier Pure-based CRIS system, metadata workflows and validation, integration with SAP and ORCID, and the use of OpenAlex.
  • 2026-07-13 to 2026-07-30 – Correspondence with representatives of Amsterdam institutions continued, yielding materials on repositories, CRIS systems, digital curation, knowledge graphs, and AI applications in library services.
  • 2026-07-26 – A follow-up round of correspondence was conducted with representatives of the Data Science Centre, Amsterdam University Library, International Institute of Social History, Information Retrieval Lab, UvA AI Chat, Openbare Bibliotheek Amsterdam, and other units.
  • 2026-07-27 – Prof. Paul Groth provided information and access to materials concerning SURF Open Research Information and agent-based tools. Analysis and testing began with a view to using the solution as a model for the experiment planned at the Jagiellonian Library.
  • 2026-07-30 – Full responses and documentation were received from Robert Gillesse of the International Institute of Social History; the main stage of correspondence with Prof. Paul Groth was also completed.
  • 2026-08-05 – Contact was established with the UvA AI Chat/ICTS team; the research questions were forwarded to specialists responsible for developing and operating the platform.
  • 2026-08-07 – The data sprint report was published: AI Dreams of Russian Propaganda, Serge Poliakoff, Julia Kling, Władysław M. Kolasa, Milena Gajewska, Weronika Gorajczyk, Piotr Bukański, Julia Puchkov, and Silvio Fernando // Digital Methods Summer School and Data Sprint 2026, University of Amsterdam – https://www.digitalmethods.net/Dmi/SummerSchool2026AIDreamsofRussianPropaganda
  • 2026-09-01 to 2026-11-15 – Planned second stage of the internship at the Jagiellonian Library, including the LLM application experiment, synthesis of findings, and preparation of the report and scholarly publication.

Last updated: 2026-08-09

 

Scroll to Top
Skip to content