{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,12,31]],"date-time":"2025-12-31T00:18:12Z","timestamp":1767140292470,"version":"build-2238731810"},"reference-count":39,"publisher":"Springer Science and Business Media LLC","issue":"29","license":[{"start":{"date-parts":[[2025,9,5]],"date-time":"2025-09-05T00:00:00Z","timestamp":1757030400000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2025,9,5]],"date-time":"2025-09-05T00:00:00Z","timestamp":1757030400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["Neural Comput &amp; Applic"],"published-print":{"date-parts":[[2025,10]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>The utilization of large language models (LLMs) in research is becoming increasingly prevalent, as they offer advanced capabilities in processing and generating human-like text. However, this advancement comes with a significant trade-off in terms of time and computational costs. In this paper, we demonstrate that analyzing large text datasets with the use of LLMs can be performed efficiently in terms of both time and energy. For this purpose, we utilize the Llama pre-trained model. In more detail, we study the topic modeling task where the goal is to discover and identify topics in large text corpora. The basis of our approach is a hierarchical divisive clustering technique that clusters the data based on their semantic similarity, after employing a Sentence-BERT encoder, pre-trained on a variety of data across different tasks. Then, using an LLM, we identify topics for representative samples from each cluster. Additionally, we introduce a new evaluation method that leverages the capabilities of LLMs to assess the alignment between discovered topics and ground truth labels, providing a robust validation metric. Our findings indicate that it is possible to effectively reduce the computational cost of the topic modeling process compared to the direct application of LLMs and BERTopic, while simultaneously enhancing inference time and overall efficiency, thereby surpassing the current state-of-the-art capabilities of BERTopic.<\/jats:p>","DOI":"10.1007\/s00521-025-11593-9","type":"journal-article","created":{"date-parts":[[2025,9,5]],"date-time":"2025-09-05T13:16:18Z","timestamp":1757078178000},"page":"24421-24439","update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["Large language models for efficient topic modeling"],"prefix":"10.1007","volume":"37","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-8607-5518","authenticated-orcid":false,"given":"Panagiotis C.","family":"Theocharopoulos","sequence":"first","affiliation":[]},{"given":"Panagiotis","family":"Anagnostou","sequence":"additional","affiliation":[]},{"given":"Spiros V.","family":"Georgakopoulos","sequence":"additional","affiliation":[]},{"given":"Sotiris K.","family":"Tasoulis","sequence":"additional","affiliation":[]},{"given":"Vassilis P.","family":"Plagianakos","sequence":"additional","affiliation":[]}],"member":"297","published-online":{"date-parts":[[2025,9,5]]},"reference":[{"key":"11593_CR1","doi-asserted-by":"publisher","unstructured":"Barde BV, Bainwad AM (2017) An overview of topic modeling methods and tools. In: 2017 International conference on intelligent computing and control systems (ICICCS), pp 745\u2013750. https:\/\/doi.org\/10.1109\/ICCONS.2017.8250563","DOI":"10.1109\/ICCONS.2017.8250563"},{"issue":"Jan","key":"11593_CR2","first-page":"993","volume":"3","author":"DM Blei","year":"2003","unstructured":"Blei DM, Ng AY, Jordan MI (2003) Latent Dirichlet allocation. J Mach Learn Res 3(Jan):993\u20131022","journal-title":"J Mach Learn Res"},{"key":"11593_CR3","doi-asserted-by":"publisher","first-page":"102131","DOI":"10.1016\/j.is.2022.102131","volume":"112","author":"A Abdelrazek","year":"2023","unstructured":"Abdelrazek A, Eid Y, Gawish E, Medhat W, Hassan A (2023) Topic modeling algorithms and applications: a survey. Inf Syst 112:102131. https:\/\/doi.org\/10.1016\/j.is.2022.102131","journal-title":"Inf Syst"},{"key":"11593_CR4","doi-asserted-by":"publisher","DOI":"10.1016\/j.is.2020.101582","volume":"94","author":"I Vayansky","year":"2020","unstructured":"Vayansky I (2020) Kumar SAP: a review of topic modeling methods. Inf Syst 94:101582. https:\/\/doi.org\/10.1016\/j.is.2020.101582","journal-title":"Inf Syst"},{"key":"11593_CR5","unstructured":"Vaswani A, Shazeer N, Parmar N, Uszkoreit J, Jones L, Gomez AN, Kaiser L, Polosukhin I (2017) Attention is all you need (nips), 10:0140525\u201316001837. arXiv:1706.03762"},{"key":"11593_CR6","doi-asserted-by":"publisher","DOI":"10.1016\/j.lindif.2023.102274","volume":"103","author":"E Kasneci","year":"2023","unstructured":"...Kasneci E, Se\u00dfler K, K\u00fcchemann S, Bannert M, Dementieva D, Fischer F, Gasser U, Groh GL, G\u00fcnnemann S, H\u00fcllermeier E, Krusche S, Kutyniok G, Michaeli T, Nerdel C, Pfeffer J, Poquet O, Sailer M, Schmidt A, Seidel T, Stadler M, Weller J, Kuhn J, Kasneci G (2023) Chatgpt for good? On opportunities and challenges of large language models for education. Learn Individ Differ 103:102274","journal-title":"Learn Individ Differ"},{"issue":"6","key":"11593_CR7","doi-asserted-by":"publisher","first-page":"3346","DOI":"10.1002\/mde.3882","volume":"44","author":"K Wang","year":"2023","unstructured":"Wang K, Lu L, Fang J, Xing Y, Tong Z, Wang L (2023) The downside of artificial intelligence (AI) in green choices: how AI recommender systems decrease green consumption. Manag Decis Econ 44(6):3346\u20133353","journal-title":"Manag Decis Econ"},{"key":"11593_CR8","unstructured":"Nuclear Energy Agency (2023) COP28 Recognises the critical role of nuclear energy for reducing the effects of climate change. https:\/\/www.oecd-nea.org\/jcms\/pl_89153\/cop28-recognises-the-critical-role-of-nuclear-energy-for-reducing-the-effects-of-climate-change"},{"key":"11593_CR9","unstructured":"European Commission (2023) Energy Efficiency Targets. https:\/\/energy.ec.europa.eu\/topics\/energy-efficiency\/energy-efficiency-targets-directive-and-rules\/energy-efficiency-targets_en"},{"issue":"9","key":"11593_CR10","doi-asserted-by":"publisher","first-page":"3464","DOI":"10.1021\/acs.est.3c01106","volume":"57","author":"MC Rillig","year":"2023","unstructured":"Rillig MC, \u00c5gerstrand M, Bi M, Gould KA, Sauerland U (2023) Risks and benefits of large language models for the environment. Environ Sci Technol 57(9):3464\u20133466","journal-title":"Environ Sci Technol"},{"key":"11593_CR11","doi-asserted-by":"crossref","unstructured":"Strubell E, Ganesh A, McCallum A (2019) Energy and policy considerations for deep learning in NLP. arXiv:1906.02243","DOI":"10.18653\/v1\/P19-1355"},{"issue":"13","key":"11593_CR12","doi-asserted-by":"publisher","first-page":"5645","DOI":"10.1016\/j.eswa.2015.02.055","volume":"42","author":"D O\u2019Callaghan","year":"2015","unstructured":"O\u2019Callaghan D, Greene D, Carthy J, Cunningham P (2015) An analysis of the coherence of descriptors in topic modeling. Expert Syst Appl 42(13):5645\u20135657. https:\/\/doi.org\/10.1016\/j.eswa.2015.02.055","journal-title":"Expert Syst Appl"},{"issue":"12","key":"11593_CR13","doi-asserted-by":"publisher","first-page":"2928","DOI":"10.1109\/TKDE.2014.2313872","volume":"26","author":"X Cheng","year":"2014","unstructured":"Cheng X, Yan X, Lan Y, Guo J (2014) BTM: topic modeling over short texts. IEEE Trans Knowl Data Eng 26(12):2928\u20132941. https:\/\/doi.org\/10.1109\/TKDE.2014.2313872","journal-title":"IEEE Trans Knowl Data Eng"},{"key":"11593_CR14","doi-asserted-by":"publisher","DOI":"10.1016\/j.neucom.2024.128941","volume":"616","author":"S Zhang","year":"2025","unstructured":"Zhang S, Xu J, Hui N, Zhai P (2025) A short text topic modeling method based on integrating Gaussian and Logistic coding networks with pre-trained word embeddings. Neurocomputing 616:128941. https:\/\/doi.org\/10.1016\/j.neucom.2024.128941","journal-title":"Neurocomputing"},{"key":"11593_CR15","unstructured":"Grootendorst, M (2022) Bertopic: Neural topic modeling with a class-based TF-IDF procedure. arXiv:2203.05794"},{"key":"11593_CR16","doi-asserted-by":"publisher","DOI":"10.1016\/j.eswa.2024.124850","volume":"256","author":"R Wang","year":"2024","unstructured":"Wang R, Wang Y, Liu X, Huang H, Sun G (2024) Bridging spherical mixture distributions and word semantic knowledge for Neural Topic Modeling. Expert Syst Appl 256:124850. https:\/\/doi.org\/10.1016\/j.eswa.2024.124850","journal-title":"Expert Syst Appl"},{"issue":"10","key":"11593_CR17","doi-asserted-by":"publisher","first-page":"10616","DOI":"10.1109\/TKDE.2023.3267496","volume":"35","author":"Z Lei","year":"2023","unstructured":"Lei Z, Liu H, Yan J, Rao Y, Li Q (2023) NMTF-LTM: towards an alignment of semantics for lifelong topic modeling. IEEE Trans Knowl Data Eng 35(10):10616\u201310632. https:\/\/doi.org\/10.1109\/TKDE.2023.3267496","journal-title":"IEEE Trans Knowl Data Eng"},{"key":"11593_CR18","doi-asserted-by":"publisher","unstructured":"Mu Y, Dong C, Bontcheva K, Song X (2024) Large language models offer an alternative to the traditional approach of topic modelling. https:\/\/doi.org\/10.48550\/arXiv.2403.16248","DOI":"10.48550\/arXiv.2403.16248"},{"key":"11593_CR19","doi-asserted-by":"publisher","unstructured":"Pham CM, Hoyle A, Sun S, Resnik P, Iyyer M (2024) TopicGPT: a Prompt-based topic modeling framework. https:\/\/doi.org\/10.48550\/arXiv.2311.01449","DOI":"10.48550\/arXiv.2311.01449"},{"key":"11593_CR20","doi-asserted-by":"publisher","unstructured":"Xu W, Hu W, Wu F, Sengamedu S (2023) DeTiME: diffusion-Enhanced Topic Modeling using Encoder-decoder based LLM. In: Findings of the association for computational linguistics: EMNLP 2023, pp 9040\u20139057. https:\/\/doi.org\/10.18653\/v1\/2023.findings-emnlp.606","DOI":"10.18653\/v1\/2023.findings-emnlp.606"},{"issue":"3","key":"11593_CR21","doi-asserted-by":"publisher","first-page":"1328","DOI":"10.1109\/TKDE.2023.3303465","volume":"36","author":"DC Zhang","year":"2024","unstructured":"Zhang DC, Lauw HW (2024) Topic modeling on document networks with Dirichlet optimal transport barycenter. IEEE Trans Knowl Data Eng 36(3):1328\u20131340. https:\/\/doi.org\/10.1109\/TKDE.2023.3303465","journal-title":"IEEE Trans Knowl Data Eng"},{"key":"11593_CR22","doi-asserted-by":"publisher","DOI":"10.1016\/j.eswa.2024.124631","volume":"255","author":"Z Luo","year":"2024","unstructured":"Luo Z, Liu L, Ananiadou S, Xie Q (2024) Graph contrastive topic model. Expert Syst Appl 255:124631. https:\/\/doi.org\/10.1016\/j.eswa.2024.124631","journal-title":"Expert Syst Appl"},{"issue":"1","key":"11593_CR23","doi-asserted-by":"publisher","first-page":"351","DOI":"10.1109\/TKDE.2024.3487523","volume":"37","author":"Y Yang","year":"2025","unstructured":"Yang Y, Lalor JP, Abbasi A, Zeng DD (2025) Hierarchical Deep Document Model. IEEE Trans Knowl Data Eng 37(1):351\u2013364. https:\/\/doi.org\/10.1109\/TKDE.2024.3487523","journal-title":"IEEE Trans Knowl Data Eng"},{"key":"11593_CR24","doi-asserted-by":"publisher","DOI":"10.1016\/j.eswa.2020.113709","volume":"161","author":"M Belford","year":"2020","unstructured":"Belford M, Greene D (2020) Ensemble topic modeling using weighted term co-associations. Expert Syst Appl 161:113709. https:\/\/doi.org\/10.1016\/j.eswa.2020.113709","journal-title":"Expert Syst Appl"},{"issue":"84","key":"11593_CR25","doi-asserted-by":"publisher","first-page":"5024","DOI":"10.21105\/joss.05024","volume":"8","author":"P Anagnostou","year":"2023","unstructured":"Anagnostou P, Tasoulis S, Plagianakos VP, Tasoulis D (2023) Hipart: hierarchical divisive clustering toolbox. J Open Source Softw 8(84):5024. https:\/\/doi.org\/10.21105\/joss.05024","journal-title":"J Open Source Softw"},{"key":"11593_CR26","doi-asserted-by":"publisher","unstructured":"Ethayarajh K (2019) How contextual are contextualized word representations? Comparing the geometry of BERT, ELMo, and GPT-2 Embeddings. https:\/\/doi.org\/10.48550\/arXiv:1909.00512","DOI":"10.48550\/arXiv:1909.00512"},{"key":"11593_CR27","doi-asserted-by":"crossref","unstructured":"Reimers N, Gurevych I (2019) Sentence-BERT: sentence embeddings using siamese BERT-Networks. In: Proceedings of the 2019 conference on empirical methods in natural language processing, pp 3982\u20133992. Association for Computational Linguistics","DOI":"10.18653\/v1\/D19-1410"},{"key":"11593_CR28","doi-asserted-by":"publisher","first-page":"325","DOI":"10.1023\/A:1009740529316","volume":"2","author":"D Boley","year":"1998","unstructured":"Boley D (1998) Principal direction divisive partitioning. Data Min Knowl Disc 2:325\u2013344","journal-title":"Data Min Knowl Disc"},{"issue":"10","key":"11593_CR29","doi-asserted-by":"publisher","first-page":"3391","DOI":"10.1016\/j.patcog.2010.05.025","volume":"43","author":"SK Tasoulis","year":"2010","unstructured":"Tasoulis SK, Tasoulis DK, Plagianakos VP (2010) Enhancing principal direction divisive clustering. Pattern Recogn 43(10):3391\u20133411","journal-title":"Pattern Recogn"},{"key":"11593_CR30","doi-asserted-by":"crossref","unstructured":"Nguyen TT, Wilson C, Dalins J (2023) Fine-tuning llama 2 large language models for detecting online sexual predatory chats and abusive texts. arXiv:2308.14683. arxiv:2308.14683","DOI":"10.14428\/esann\/2024.ES2024-222"},{"key":"11593_CR31","unstructured":"Dubey A, Jauhri A, Pandey A, Kadian A, Al-Dahle A, Letman A, Mathur A, Schelten A, Yang A, Fan A, et al (2024) The llama 3 herd of models. arXiv:2407.21783"},{"key":"11593_CR32","doi-asserted-by":"publisher","unstructured":"Greene D, Cunningham P (2006) Practical solutions to the problem of diagonal dominance in kernel document clustering. In: Proceedings of the 23rd international conference on machine learning, pp 377\u2013384. https:\/\/doi.org\/10.1145\/1143844.1143892","DOI":"10.1145\/1143844.1143892"},{"key":"11593_CR33","unstructured":"Sudarshan (2021) A topic modeling for research articles"},{"key":"11593_CR34","doi-asserted-by":"publisher","unstructured":"Gautam: E Commerce Text Dataset. Zenodo (2019). https:\/\/doi.org\/10.5281\/zenodo.3355823","DOI":"10.5281\/zenodo.3355823"},{"key":"11593_CR35","doi-asserted-by":"crossref","unstructured":"Kwon W, Li Z, Zhuang S, Sheng Y, Zheng L, Yu CH, Gonzalez JE, Zhang H, Stoica I (2023) Efficient memory management for large language model serving with pagedattention. In: Proceedings of the ACM SIGOPS 29th symposium on operating systems principles","DOI":"10.1145\/3600006.3613165"},{"key":"11593_CR36","first-page":"31","volume":"30","author":"G Bouma","year":"2009","unstructured":"Bouma G (2009) Normalized (pointwise) mutual information in collocation extraction. Proc GSCL 30:31\u201340","journal-title":"Proc GSCL"},{"key":"11593_CR37","doi-asserted-by":"publisher","first-page":"439","DOI":"10.1162\/tacl_a_00325","volume":"8","author":"AB Dieng","year":"2020","unstructured":"Dieng AB, Ruiz FJ, Blei DM (2020) Topic modeling in embedding spaces. Trans Assoc Comput Linguist 8:439\u2013453","journal-title":"Trans Assoc Comput Linguist"},{"key":"11593_CR38","doi-asserted-by":"crossref","unstructured":"Vinh N, Epps J, Bailey J (2009) Information theoretic measures for clusterings comparison: Variants. Properties, Normalization and Correction for Chance, vol 18","DOI":"10.1145\/1553374.1553511"},{"key":"11593_CR39","doi-asserted-by":"crossref","unstructured":"Kogan J, Nicholas CK, Teboulle M (eds) (2006) Grouping multidimensional data: recent advances in clustering. Springer, Berlin","DOI":"10.1007\/3-540-28349-8"}],"updated-by":[{"DOI":"10.1007\/s00521-025-11691-8","type":"correction","label":"Correction","source":"publisher","updated":{"date-parts":[[2025,11,17]],"date-time":"2025-11-17T00:00:00Z","timestamp":1763337600000}}],"container-title":["Neural Computing and Applications"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s00521-025-11593-9.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/article\/10.1007\/s00521-025-11593-9\/fulltext.html","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s00521-025-11593-9.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,11,17]],"date-time":"2025-11-17T06:58:49Z","timestamp":1763362729000},"score":1,"resource":{"primary":{"URL":"https:\/\/link.springer.com\/10.1007\/s00521-025-11593-9"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,9,5]]},"references-count":39,"journal-issue":{"issue":"29","published-print":{"date-parts":[[2025,10]]}},"alternative-id":["11593"],"URL":"https:\/\/doi.org\/10.1007\/s00521-025-11593-9","relation":{},"ISSN":["0941-0643","1433-3058"],"issn-type":[{"value":"0941-0643","type":"print"},{"value":"1433-3058","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,9,5]]},"assertion":[{"value":"23 January 2025","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"12 August 2025","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"5 September 2025","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"2 November 2025","order":5,"name":"change_date","label":"Change Date","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"Update","order":6,"name":"change_type","label":"Change Type","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"The original online version of this article was revised due to Retrospective Open Access order","order":7,"name":"change_details","label":"Change Details","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"17 November 2025","order":8,"name":"change_date","label":"Change Date","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"Correction","order":9,"name":"change_type","label":"Change Type","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"A Correction to this paper has been published:","order":10,"name":"change_details","label":"Change Details","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"https:\/\/doi.org\/10.1007\/s00521-025-11691-8","URL":"https:\/\/doi.org\/10.1007\/s00521-025-11691-8","order":11,"name":"change_details","label":"Change Details","group":{"name":"ArticleHistory","label":"Article History"}},{"order":1,"name":"Ethics","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.","order":2,"name":"Ethics","group":{"name":"EthicsHeading","label":"Conflict of interest"}}]}}