The advancement of artificial intelligence (AI) shows the potential to facilitate information organisation, providing opportunities for classification applications in libraries. While previous studies have investigated the potential usage of AI in libraries, its real-world integration into specific tasks like cataloguing is limited. In our study, we employ large language models (LLMs) to develop classifiers for book classification using the Library of Congress Classification system. We experiment with different input data, data sizes and models to identify effective settings for assisting cataloguing using LLMs. In addition, we evaluate the model performance at different levels of Library of Congress Classification (LCC) system class granularity. The results show that Llama3 outperforms the other five models in our experiment. More diverse input data may contribute to enhancing model performance. However, LLMs demonstrate limitations in processing long text and handling more granular categories.
ClarkeRISchoonmakerS. The critical catalog: giving voice to diverse library materials through provocative design. Adv Classif Res Online2018: 23–26.
2.
RaiaanMAKMuktaMSHFatemaK, et al. A review on large language models: architectures, applications, taxonomies, open issues and challenges. IEEE Access2024; 12: 26839–26874.
3.
ChowEHKaoTJLiX. An experiment with the use of ChatGPT for LCSH subject assignment on electronic theses and dissertations. Catalog Classif Q2024; 62(5): 574–588.
BasyalLSanghviM. Text summarization using large language models: a comparative study of MPT-7B-instruct, Falcon-7B-instruct, and OpenAI Chat-GPT models. arXiv preprint, arXiv:2310.10449, 2023.
9.
ZhaoXOuyangSYuZ, et al. Pre-trained language models can be fully zero-shot learners. In: RogersABoyd-GraberJOkazakiN (eds) Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)Toronto, ON: Association for Computational Linguistics, 2023, pp. 15590–15606.
10.
SunXLiXLiJ, et al. Text classification via large language models. In: BouamorHPinoJBaliK (eds) Findings of the association for computational linguistics: EMNLP. Kerrville, TX: Association for Computational Linguistics, 2023, pp. 8990–9005.
11.
XieQBishopJATiwariP, et al. Pre-trained language models with domain knowledge for biomedical extractive summarization. Knowl Based Syst2022; 252: 109460.
12.
KrageljMKljajić BorštnarM. Automatic classification of older electronic texts into the Universal Decimal Classification–UDC. J Doc2021; 77(3): 755–776.
13.
LyuLHanT. A comparative study of Chinese patent literature automatic classification based on deep learning. In: 2019 ACM/IEEE joint conference on digital libraries (JCDL), Champaign, IL, 2–6 June 2019. New York: IEEE, pp. 345–346.
14.
YuHF. Bibliographic automatic classification algorithm based on semantic space transformation. Multimed Tools Appl2020; 79: 9283–9297.
15.
ZhangWDengYLiuB, et al. Sentiment analysis in the era of large language models: a reality check. In: DuhKGomezHBethardS (eds) Findings of the association for computational linguistics: NAACL. Kerrville, TX: Association for Computational Linguistics, 2024, pp. 3881–3906.
16.
ZhangYWangMRenC, et al. Pushing the limit of LLM capacity for text classification. arXiv preprint, arXiv:2402.07470, 2024.
17.
DagdelenJDunnALeeS, et al. Structured information extraction from scientific text with large language models. Nat Commun2024; 15(1): 1418.
18.
BrownTMannBRyderN, et al. Language models are few-shot learners. Adv Neural Inf Process Syst2020; 33: 1877–1901.
19.
HowardJRuderS. Universal language model fine-tuning for text classification. In: GurevychIMiyaoY (eds) Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics. (Volume 1: Long Papers). Kerrville, TX: Association for Computational Linguistics, 2018, pp. 328–339.
20.
KojimaTGuSSReidM, et al. Large language models are zero-shot reasoners. Adv Neural Inf Process Syst2022; 35: 22199–22213.
21.
HuEJWallisPAllen-ZhuZ, et al. LoRA: low-rank adaptation of large language models. International Conference on Learning Representations, 2021, https://github.com/microsoft/LoRA
22.
WettigA.GuptaA.MalikS.ChenD. (2024, July). QuRating: selecting high-quality data for training language models. In Proceedings of the 41st International Conference on Machine Learning (pp. 52915–52971).
23.
ZhuXFuYZhouB, et al. Critical data size of language models from a Grokking perspective. arXiv preprint, arXiv:2401.10463, 2024.
24.
Dos SantosVGSantosGLLynnT, et al. Identifying citizen-related issues from social media using LLM-based data augmentation. In: GuizzardiGSantoroFMouratidisH, et al. (eds) International conference on advanced information systems engineering. Cham: Springer Nature, 2024, pp. 531–546.
25.
LiYLvZ. BookGCN: a book classification model based on text attributes and link relationships. In: International conference on cloud computing and communication engineering (CCCE 2024), Nanjing, China, 14–16 June 2024. Bellingham, WA: SPIE.
26.
GolubKHagelbäckJArdöA. Automatic classification of Swedish metadata using Dewey Decimal Classification: a comparison of approaches. J Data Inf Sci2020; 5(1): 18–38.
27.
Van der BurghBVerberneS. The merits of universal language model fine-tuning for small datasets – a case with Dutch book reviews. arXiv preprint, arXiv:1910.00896, 2019.
28.
AbdelrahmanEFoxE. Improving accessibility to Arabic ETDs using automatic classification. In: 26th international conference on theory and practice of digital libraries, TPDL 2022, Padua, 20–23 September 2022, pp. 230–242. New York: Springer.
29.
FormanekM. Exploring the potential of large language models and generative artificial intelligence (GPT): applications in library and information science. J Librariansh Inf Sci2025; 57(2): 568–590.
30.
BagchiM. Toward generative AI–Driven metadata modeling: a human–large Language model collaborative approach. Libr Trends2025; 73(3): 297–322.
31.
BuschLTebernumDVelardeG. Exploring LLM capabilities in extracting DCAT-compatible metadata for data cataloging. arXiv preprint, arXiv:2507.05282, 2025.
32.
LuoPLiuZ. Automatic subject indexing with large language model fine-tuning and controlled vocabulary retrieval. Aslib J Inf Manag2025; 2025: 1–20.
33.
BorovičMTomovskiELi DobnikT, et al. Evaluating proprietary and open-weight large language models as Universal Decimal Classification recommender systems. Appl Sci2025; 15(14): 7666.
34.
ZhangYYangRXuX, et al. Teleclass: taxonomy enrichment and LLM-enhanced hierarchical text classification with minimal supervision. arXiv preprint, arXiv:2403.00165, 2024.
35.
MartinsS. Artificial intelligence-assisted classification of library resources: the case of claude AI. Library Philosophy & Practice.
36.
CarvalhoDVPereiraEMCardosoJS. Machine learning interpretability: a survey on methods and metrics. Electronics2019; 8(8): 832.
HeJRungtaMKoleczekD, et al. Does prompt formatting have any impact on LLM performance? arXiv preprint, arXiv:2411.10541, 2024.
45.
KeithCRobinsonMDuncanF, et al. Optimizing large language models: a novel approach through dynamic token pruning, 2024.
46.
WangCDuanHZhangS, et al. Ada-LEval: evaluating long-context LLMs with length-adaptable benchmarks. In: DuhKGomezHBethardS (eds) Proceedings of the 2024 conference of the North American chapter of the association for computational linguistics: human language technologies (Volume 1: Long Papers). Kerrville, TX: Association for Computational Linguistics, 2024, pp. 3712–3724.
47.
WilliamsCK. The effect of class imbalance on precision-recall curves. Neural Comput2021; 33(4): 853–857.
48.
LiuJSongXZhangD, et al. A hybrid framework for subject analysis: integrating embedding-based regression models with large language models. Proc Assoc Inf Sci Technol2025; 62(1): 445–457.
49.
GaoTJinJKeZT, et al. A comparison of DeepSeek and other LLMs. Am Stat2026; 80: 164–176.
50.
WangCYouFWangY. Text intelligent classification model based on improved KNN algorithm in library service transformation. Syst Soft Comput2025; 7: 200313.
MichalakREllixsonD. Reimagining libraries in the AI era: ethical AI literacy and human-centered strategies. AI Soc. Epub ahead of print 7January2026. DOI: 10.1007/s00146-025-02827-4.
53.
ElroukhSM. Collaborative cataloging with AI: an experimental study on OpenAI’s ChatGPT-3.5. Int J Comput Sci Netw Secur2025; 25(6): 159–167.
54.
KlairHGermanFBanerjeeB, et al. Evaluating human-LLM alignment in ETD subject classification. In: BalkeWTGolubKManolopoulosY, et al. (eds) New trends in theory and practice of digital libraries (TPDL 2025. Communications in Computer and Information Science), Vol. 2694. Cham: Springer, 2026. DOI: 10.1007/978-3-032-06136-2_6.
55.
HindMHoudeSMartinoJ, et al. Experiences with improving the transparency of AI models and services. In: Extended Abstracts of the 2020 CHI Conference on Human Factors in Computing Systems, Honolulu, HI, 25–30 April 2020, pp. 1–8. New York: ACM. DOI: 10.1145/3334480.3383051.
