Search

Talk at the Huminfra seminar "AI and the Library: Subject Descriptions in the Era of Artificial Intelligence", Borås, 29 January 2024

Linnaeus University (Linnéuniversitetet, LNU) is an international public university in the province of Småland, Sweden. LNU was founded in 2010 by a merger of the former Växjö University and Kalmar University, and currently has approximately 44,000 enrolled students. The university is currently Sweden’s sixth largest in terms of student numbers. It has 600 partner universities in more than 80 countries around the world. Over the past decade, there has been a distinct emphasis on the Digital Humanities at LNU through a variety of initiatives, all focused on fostering interdisciplinary expertise in the Humanities, data analysis, cultural heritage, and ICTs. Best described as a decentralized collaborative culture, DH at LNU includes knowledge environments (e.g. Digital Transformations), centers of excellence (e.g. the Centre for Data Intensive Sciences and Applications), and the iInstitute (the local center for the international iSchools consortium). LNU was the first Swedish university to join DARIAH and is now leading the bid for national membership. In 2016, Linnaeus established a Digital Humanities Hub to focus on data-intensive digital humanities, leading to the implementation of digital humanities as a research and teaching subject at the Faculty of Arts and Humanities, and to the Digital Humanities MA programme. This programme is being offered in English to international students, who benefit from being able to take advantage of the worldwide iSchools agreement for virtual student and faculty exchange. As part of the work of the DH Hub and the iInstitute, LNU was recently granted funding for a national PhD school in digital humanities, an initiative between four Swedish universities. In this presentation we will outline the original vision for fostering DH at Linnaeus University, reflect on the challenges and successes of the past few years and present general ideas on how to facilitate DH at the intersection of multiple disciplines.

Subject indexing of LGBTQ+ (lesbian, gay, bi, trans, queer, and ace) materials is often criticized for being too general and superficial, as general subject headings systems rarely provide the desirable depth and breadth. LGBTQ+ fiction is an even bigger problem because fiction is most often indexed only for genre, place, and time, while themes remain unaddressed. Thus, many readers look outside the library catalogue to identify LGBTQ+ titles in social medial, personal social networks, or web search engines. This exploratory study builds on the Queerlit database of over 1800 Swedish LGBTQ+ works of fiction which have been subject indexed using a dedicated thesaurus. It aims to identify and discuss how Queerlit subject terms compare with those in the Swedish Union Catalogue (Libris), the social cataloguing website Goodreads as well as the Google Books API. In addition, subject access to LGBTQ+ works of fiction in China is discussed, particularly via Douban, the social networking platform.

This article addresses the underutilization of knowl­edge organization systems (KOS) elements in online library catalogs, hindering effective subject-based search and discovery. It highlights the International Society for Knowl­edge Organization's initiative to develop metadata guidelines for library catalog procurement, focusing on maximizing the value of subject metadata from classification systems and controlled vocabularies. The paper discusses the rationale for quality subject access, proposes desirable search functionalities based on research, explores implementation challenges, and outlines future developments. The conclusion emphasizes the importance of providing quality subject access in digital services and calls for further research on interface design, guideline adoption, KOS evolution, and the impact of language models on subject metadata use. The work underscores the need for applying controlled vocabularies in search interfaces across libraries, archives, and museums while acknowledging the complementary role of alternative approaches like social tagging and automatic indexing. Extensive future research is suggested to implement search functionalities, promote guidelines adoption, enhance KOS evolution, and assess the influence of language models on subject metadata utilization.

We are currently in a phase where it seems that new applications for large language models (LLMs) in general and generative pre-trained transformers (GPTs) in particular are tested every day. Examples are as diverse as automated data mining for building energy management (C. Zhang et al., 2024), evaluating the accuracy of differential-diagnosis lists for clinical vignettes (Hirosawa et al., 2023), and human-machine-augmented intelligent vehicles (J. Zhang et al., 2023). They can also be used to extract structured information from unstructured text (Söderström, 2023). This poster presents a pilot study on one such application, the potential use of OpenAI’s ChatGPT for automatic subject indexing of archival documents in Swedish, Swedish LGBTQ fiction and Chinese fiction. The accuracy of the assigned subject index terms is compared with the output from ANNIF (Suominen et al., 2022), an established automatic subject indexing software used in libraries. The results display an impressive degree of accuracy for the subject index terms assigned by ChatGPT, but challenges have been identified in all three document types. For example, the appropriateness of the terms for historical text is highly questionable at times. The terms assigned by ANNIF, in contrast, are drawn from a controlled vocabulary, which ensures that they have been manually selected as suitable subject index terms. The pilot study shows that it is feasible to run the index terms suggested by ChatGPT through ANNIF to get index terms from a controlled vocabulary while harnessing ChatGPT’s state-of-the-art natural language understanding. This presents intriguing opportunities for implementing GPTs in the archival/library cataloging workflows. Semi-automatic approaches and manual checks are still to be preferred, however, in order to maintain the authenticity of the generated metadata.

Has something been lost in the move towards federated searching and relevance ranking in academic library search interfaces? The richness of metadata and other knowledge organisation systems (KOSs) such as classification schemes and controlled vocabularies is often not featured in library management system (LMS) procurement. The ISKO Scientific and Technical Advisory Committee (STAC) has set up a working group to develop a set of guidelines for academic and research libraries that are planning to procure an LMS. The purpose of the guidelines is to enable users to benefit from the intellectual effort that has been invested in assigning metadata to electronic resources. This paper reports on progress of the working group in its review of the literature and the plans for consultations with stakeholders.

Invited panel talk at DARIAH Day for Iceland, 31 May 2024Ej belagd 240624

DHNB Workshop on Higher Education in DH, Reykjavik, May 2024

Huminfra

Ej belagd 231213

Dimpah project

Humanities scholars can today engage in research inquiry using data from a range of varied collections which are often characterised by poor subject access, often resulting in systems that underperform and even effectively prevent access to data, information and knowledge. In spite of the availability of professional standards and guidelines to provide quality-controlled subject access through knowledge organisation systems (KOS), subject access in such collections is rarely based on KOS. At the same time, KOS themselves may come with problems such as being slow to update, being rigidly structured and not incorporating end-users’ vocabulary. It may therefore be useful to consider methods for remediating these deficiencies in KOSs, such as collecting user-generated metadata via social tagging or complementing automated indexing techniques with manual ones. To help address the above problems, the paper discusses these challenges and points to possible solutions in different contexts. It does so by reflecting on an open educational resource (OER) devoted to this theme, titled Introduction to Knowledge Organisation Systems for Digital Humanities. It was developed as part of an EU project called DiMPAH (Digital Methods Platform for the Arts and Humanities), 2021–2023, creating seven OERs for inclusion in DARIAH Teach., Erasmus+ Dimpah project