Automated subject indexing in large-scale library collections: a minimal semantic enhancement strategy at the National Library of Spain
Author:
- 26 August 2026
- journal article
- Published by Emerald Publishing in Library Hi Tech
- Vol. 44 (4-5) , 728-745
Abstract
Purpose: This study addresses a key technical challenge in automated subject indexing: improving prediction accuracy for large-scale library collections. We introduce and validate a hybrid AI framework that combines statistical and semantic embedding approaches to enhance the performance of automated subject heading assignment. The primary aim is to demonstrate how a minimal, targeted application of a powerful semantic embedding model can significantly boost the accuracy of a high-speed statistical classifier, particularly for descriptive, context-rich bibliographic titles. Design/methodology/approach: A large-scale validation was performed using 2.4 million real-world bibliographic records from the National Library of Spain. We developed a hybrid framework combining a high-speed statistical engine (Omikuji) with an advanced semantic model (Qwen3-Embedding-8B). A key finding emerged during validation: an optimal “97/3” resource-efficient blend, where the statistical model handles 97% of the weight, and the semantic embedding model provides a minimal 3% corrective signal. The system's prediction accuracy was then evaluated on distinct test sets, including three stratified by title length (short, medium and long) to assess performance across varying levels of contextual richness. Findings: The results demonstrate significant accuracy improvements. First, the “less is more” hypothesis is confirmed: a minimal 3% semantic contribution proves optimal, substantially improving baseline performance. Second, the enhancement effect correlates strongly with title length and contextual richness. While the hybrid model achieved a 7.3% accuracy gain (nDCG@5) on general records, performance improvements were more pronounced for longer titles: +24.9% for medium-length titles and +40.1% for long, descriptive titles, precisely the materials where statistical models typically struggle most. Originality/value: This study provides empirical evidence for an effective method of integrating semantic embedding and statistical approaches in automated subject indexing. Its contribution lies in demonstrating that substantial accuracy gains can be achieved through minimal semantic enhancement rather than complete system replacement. By systematically evaluating performance across different title complexities, this research offers practical guidance for institutions seeking to improve their automated subject indexing systems, particularly for descriptive bibliographic records that challenge purely statistical approaches.Keywords
This publication has 16 references indexed in Scilit:
- The Automation of Subject Indexing at ZBW and the Role of Metadata in Times of Large Language ModelsProcedia Computer Science, 2024
- AI in automating library cataloging and classificationLibrary Hi Tech News, 2024
- Automated Dewey Decimal Classification of Swedish library metadata using Annif softwareJournal of Documentation, 2024
- An Analysis of BERT (NLP) for Assisted Subject Indexing for Project GutenbergCataloging & Classification Quarterly, 2022
- Automated Subject Indexing: An OverviewCataloging & Classification Quarterly, 2021
- Automated subject indexing using word embeddings and controlled vocabularies: a comparative studyInternational Journal of Metadata, Semantics and Ontologies, 2021
- Bonsai: diverse and shallow trees for extreme multi-label classificationMachine Learning, 2020
- Toward effective automated weighted subject indexing: A comparison of different approaches in different environmentsJournal of the Association for Information Science and Technology, 2017
- Deep Learning for Extreme Multi-label Text Classification2017
- A framework for evaluating automatic indexing or classification in the context of retrievalJournal of the Association for Information Science and Technology, 2015