Advances in Linked Data and KOS Research
Published: March14, 2023 by Dr. Lei Zhang
Linked data is about publishing, sharing, and connecting data from different sources on the web. The application of linked data covers domains such as geography, government, life sciences, linguistics, media, etc. Linked data in cultural heritage institutions include cases of bibliographic data, authority data, controlled vocabularies, and metadata element sets, from national libraries, research libraries, public libraries to archives and museums. The development of conceptual model IFLA LRM and the cataloging standard RDA as well as the bibliographic framework BIBFRAME aim to enable the metadata to be more open, linked, and enriched.
KOS (Knowledge Organization Systems), referring to classification schemes, subject headings, taxonomies, thesauri, ontologies, have been increasingly published as linked data. KOS are the essential tools for information organization and retrieval and other purposes, moving toward knowledge organization structures/systems/services to be used in the networked environment.
I am currently working on a research project of the linked data use cases in libraries, archives, and museums. The purpose of this research is to delineate a picture of the components and characteristics of linked data use cases in libraries, archives, and museums to inform the future linked data policies and practices. This research is to track the ongoing development of linked data, identify the key issues and trends that are common and unique across the settings, from standards and approaches to resources and services.
The first phase study explored the barriers and challenges of linked data implementation from the perspective of knowledge management (Zhang, 2022). Through the preferred reporting items for systematic reviews and meta-analyses (PRISMA), thirty articles and reports were collected from a variety of sources over the past decade and were analyzed from multiple facets. A range of issues were identified with regard to technical, administrative, and social aspects, from data quality and reliability, inconsistency in standards and best practices to lack of resources such as tools and licenses. It suggested that the crucial role of knowledge management and innovation should deserve enough attention to empower the linked data projects and services. The next phase study is underway further examining the implications and trends of linked open data with a complete set of use cases and in a broader context.
The organization of data and information has been the foundation in information science but is transforming with the advancements of technologies and multidisciplinary knowledge and is finding its applicability to a wide variety of settings. I have delved into the evolution and the power of linked data along with emerging technologies and the potential of knowledge organization systems to fulfill in various information environments.
The development and use of linked data and KOS are covered in the two elective courses I teach, INFO 247 Vocabulary Design and INFO 287 Linked Data.
Zhang, L. (2022). Empowering linked data in cultural heritage institutions: A knowledge management perspective. Data and Information Management, 6(3). https://doi.org/10.1016/j.dim.2022.100013
Linked Data and Archives
Thanks for the CIRI post updating us on your research. I've read your article as well. I know you state libraries, archives, and museums as the institutions you include in your research, but that the initial findings relate to libraries and museums. Doing the next phase of your study when you explore use cases, are you planning to ensure that the study will include archives use cases? That is what I am specifically interested in.
I'm working on the InterPARES Trust AI research program out of UBC with four other SJSU iSchool colleagues. One the studies I am leading is on the use of paradata (procedural documentation) for AI and its relationship to metadata, provenance, and XAI. We are considering paradata as a preservation target -- one that might be used in a variety of ways -- providing an explanation for the outcomes of the AI technique, serving as evidence in court proceedings where the AI technique (tools, processes, etc.) comes into question, as the basis for risk mitigation and management when deciding on the implementation of AI solutions, and more. At some point, there are a number of artifacts being produced -- some being documented through the use of model cards, datasheets, etc. -- but some not being collected at all. At some point as our research progresses, we will need to find a way to gather the paradata and I am wondering if linked data will help us do that.
We also have teams looking at metadata in relation to AI separate from this paradata study. If you are ever interested in learning more and have the time, I can invite you to our next paradata team meeting and provide you with a few of the articles we've published on the topic.
If paradata is out of scope for now, I look forward to reading about the outcome of phase 2 of your research project.
Hi Pat,Thank you so much for
Thank you so much for your interest and response!
The findings are related to libraries, archives, and museums, but for the data collected in phase one, there were very few use cases specifically in archives. The next phase study will expand the scope with more use cases, including those of archives.
Also, thanks for the information about your research. I would be interested in learning more about paradata.
Post new comment