Report on DCMI 2024 International Conference on Dublin Core and Metadata Application

DCMI 2024 was hosted by the University of Toronto Faculty of Information from October 19-23.

The theme was “Metadata Innovation: Trust, Transformation, and Humanity.” I served on the Program Committee as the papers co-chair and was integrally involved in planning the program which included keynotes, invited panels, presentations, academic papers, tutorials, workshops, best practice showcases, a student forum, posters, and essential community updates organized into 36 sessions over 5 days. Dublin Core is a small conference that attracts metadata researchers and practitioners to local settings to intermix with and build the local community of practice.

Some of the key themes that emerged at DCMI 2024 were:

  • Teaching computational archive science, for example, generating or extracting tabular data from archival resources.
  • Applying AI to cultural heritage content, for example, automating musical notation, generating primers, building immersive experiences, or composing music from data.
  • Automating entity aggregation, “semanticization” of relationships, and metadata enrichment.

In various talks and presentations, specific examples of opportunities and practices were offered. Dublin Core applications have clearly moved beyond simple resource discovery (search) to embrace AI, particularly generative AI. Resource description continues to move towards automation. Please refer to the conference programme.

Notable were several presentations about methods for digitizing analog resources so that they can be converted into analyzable data. A tutorial on the Dublin Core Tabular Application Profile (DCTAP) was presented twice to meet the demand for information about this method. The goal is not just to digitize historical content, but to provide analyzable tabular content, and to facilitate such analyses. A day-long Dublin Core Data Institute was held to define and discuss “critical data science.” 

There was also an all-day workshop on AI and knowledge organization that included ten presentations on applications using LLMs, knowledge graphs, machine learning, and other AI methods. This workshop included an interesting presentation about the “African Knowledge Hub” a federated search of the United Nations System organized around the UN SDGs. Please see the ASKIA video (Only Tibaut has been given access to the video.) produced by Amplyfi for the UN Economic Commission for Africa.

Watch NKOS Workshop Presentations

Explore the day’s presentations below. We’ve split them up into three pieces for easier consumption and searching.

Part 1

SpeakerRecording 
Title
Joseph Busch00:0015:49Introduction
Shu-Jiun Chen15:5048:11Leveraging Generative AI for Multilingual Thesaurus Development: Insights from the Confucius Ceremony Cultural Vocabulary
Cuijuan Xi48:121:19:50The Ontology enhanced multimodal large language models for the Knowledge Organization and Representation of multi-modal cultural memory resources
Rafael Leal, Annastiina Ahola and Eero Hyvönen1:19:511:46:16Using LLMs for Enriching Metadata with Links to KOS and Knowledge Graphs: Case Finnish Named Entity Linking

Part 2

SpeakerRecording 
Title
Irene Onyancha, Ahmed Al-Awah, Fraser Gordon and Rofaida Elzubair00:0034:21The African Knowledge Hub: Role of the SDG taxonomy (KOS) in harnessing, exploring, and navigating knowledge of the United Nations System – Africa
Wei Hu, Shuying Li and Ning Yang34:2258:01Patent citation link prediction based on graph neural network
Dong Whi Yoo and Koustuv Saha58:021:35:25Exploring Patient Perspectives on Anticipating and Mitigating Potential Harms of LLMs in Depression Self-Management

Part 3

SpeakerRecording 
Title
Qiaoyi (Joy) Liu and Jian Qin00:0019:19Testing use of an LLM algorithm on Gene Ontology (GO) dataset
Margie Hlava19:2046:20Using KOS AI to expedite the conference experience
Brian Dobreski and Chris Hastings46:211:16:31Evaluating AI Assignment of Library of Congress Subject Headings (LCSH)
Trevor Watkins1:16:321:50:44Mapping the Global AI Landscape using AI and KOS
Joseph Busch1:50:452:00:00Closing Remarks