Generative AI Meets Cataloging Practice: Findings from a Comparative Pilot Study.
This study evaluates the performance of four generative AI models--ChatGPT, DeepSeek, Gemini, and Copilot--in generating descriptive metadata for bibliographic resources. Models were tested on a small, diverse set of resources using four prompt types: a basic prompt, a basic prompt with an example,...
| Publicado en: | Information Technology & Libraries Vol. 45; no. 2; pp. 1 - 21 |
|---|---|
| Autores principales: | , , |
| Formato: | computer program research tables/charts Journal Article |
| Publicado: |
American Library Association
Jun2026
|
| Acceso en línea: | Ver este registro en EBSCOhost |
| fields | @attributes: recordID: 1 pdfLink: plink: https://search.ebscohost.com/login.aspx?direct=true&db=ccm&AN=194752200&site=ehost-live header: @attributes: shortDbName: ccm uiTerm: 194752200 longDbName: CINAHL Complete uiTag: AN controlInfo: bkinfo: dissinfo: jinfo: jid: 07309295 ITL jtl: Information Technology & Libraries issn: 07309295 maglogo: N pubinfo: dt: Jun2026 vid: 45 iid: 2 pid: 55 pub: American Library Association place: Chicago, Illinois artinfo: ui: 194752200 194752200 194752200 10.5860/ital.v45i2.17499 194752200 ppf: 1 ppct: 20 formats: fmt: @attributes: type: P tig: atl: Generative AI Meets Cataloging Practice: Findings from a Comparative Pilot Study. aug: au: Heng, Greta Lampron, Patricia Han, Myung-Ja affil: Cataloging and Metadata Strategies Librarian, San Diego State University sug: subj: Artificial Intelligence, Generative Utilization Cataloging Information Management Access to Information Metadata Human Comparative Studies Pilot Studies Bibliography and References Bibliographic Control Serial Publications Abstracting and Indexing Sensitivity and Specificity Reproducibility of Results Vocabulary, Controlled Descriptive Statistics Data Collection, Computer Assisted Information Retrieval ab: This study evaluates the performance of four generative AI models--ChatGPT, DeepSeek, Gemini, and Copilot--in generating descriptive metadata for bibliographic resources. Models were tested on a small, diverse set of resources using four prompt types: a basic prompt, a basic prompt with an example, a detailed prompt referencing Resource Description and Access (RDA) guidelines, and a detailed prompt with an example. Results show that both detailed RDA guidance and the inclusion of sample outputs improved metadata quality, particularly in formatting and field structure. While DeepSeek and ChatGPT showed better performance on the tasks, all models displayed limitations in parsing and following the prompts, using descriptive metadata fields, analyzing subject headings, and assigning URIs. These findings suggest that while generative AI holds potential to assist in metadata creation, its current capabilities fall short of meeting cataloging standards without human review. pubtype: Academic Journal doctype: computer program research tables/charts Journal Article ougenre: Article language: English refInfo: holdings: @attributes: islocal: N |
|---|