{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,21]],"date-time":"2026-10-21T06:45:16Z","timestamp":1792565116701,"version":"4.5.0"},"reference-count":37,"publisher":"Emerald","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"abstract":"\n Purpose<\/jats:title>\n This study aims to examine whether large language models (LLMs) can improve tourism demand forecasting accuracy compared to established statistical and machine learning approaches.<\/jats:p>\n <\/jats:sec>\n \n Design\/methodology\/approach<\/jats:title>\n A systematic out-of-sample comparison is conducted using monthly tourism arrival data from Eurostat for eight European source markets. Benchmarks include optimized seasonal autoregressive integrated moving average (SARIMA), exponential smoothing (ETS), Prophet, Random Forest and seasonal naive models. Three LLM families are evaluated as zero-shot forecasters. Robustness is assessed across a pre-COVID stable period, a future period beyond LLM training cutoffs and an older model test.<\/jats:p>\n <\/jats:sec>\n \n Findings<\/jats:title>\n LLM-based forecasts demonstrate consistent performance advantages over all benchmark models across countries and evaluation periods. In the main post-COVID period, LLMs achieve mean MASE values of 0.67\u20130.85 compared to 1.56 for ETS, 2.67 for SARIMA, 3.04 for Prophet and 4.32 for Random Forest. The LLM advantage persists in the pre-COVID stable period (LLM MASE 0.85\u20130.90 vs SARIMA 1.71 and ETS 1.31), indicating that the benefit extends beyond structural-break contexts. Robustness tests using an older LLM model and forecasts for a future period beyond LLM training data indicate that data contamination does not explain the observed performance patterns. Prompt ablation analysis reveals that providing contextual information (country identity and date labels) is the most beneficial design choice, while seasonal and shock hints offer no consistent improvement. Diebold\u2013Mariano tests show directional evidence favoring LLMs in 31 of 32 pairwise comparisons, though limited test sample size (14\u201323 observations per country) precludes individual-country statistical significance at conventional levels.<\/jats:p>\n <\/jats:sec>\n \n Research limitations\/implications<\/jats:title>\n The analysis focuses on monthly national-level arrival data and zero-shot prompting. Future research should examine alternative frequencies, destination-level data, fine-tuned or retrieval-augmented LLM designs and hybrid approaches combining LLM outputs with statistical model adjustments. The limited test sample size (12\u201324 monthly observations per country) restricts statistical power for individual-country significance tests, though the consistency of directional advantages across all countries provides compelling qualitative evidence.<\/jats:p>\n <\/jats:sec>\n \n Practical implications<\/jats:title>\n LLMs offer a scalable decision-support tool for destination management organizations with near-zero implementation cost via API access.<\/jats:p>\n <\/jats:sec>\n \n Originality\/value<\/jats:title>\n To the best of the authors\u2019 knowledge, this study provides the first multicountry European benchmark comparing LLM-based forecasting with optimized statistical and machine learning models across multiple evaluation periods, with systematic robustness controls for data contamination.<\/jats:p>\n <\/jats:sec>","DOI":"10.1108\/tr-02-2026-0223","type":"journal-article","created":{"date-parts":[[2026,10,21]],"date-time":"2026-10-21T06:08:21Z","timestamp":1792562901000},"page":"1-20","source":"Crossref","is-referenced-by-count":0,"title":["Tourism demand forecasting in the artificial intelligence era using large language models"],"prefix":"10.1108","author":[{"ORCID":"https:\/\/orcid.org\/0000-0003-4888-6026","authenticated-orcid":true,"given":"Roman","family":"Egger","sequence":"first","affiliation":[{"name":"Modul University Vienna School of Tourism and Service Management, , Vienna,","place":["Austria"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"140","published-online":{"date-parts":[[2026,10,22]]},"reference":[{"issue":"4","key":"2026102102081514400_ref001","first-page":"1313","article-title":"Humans vs. large language models: judgmental forecasting in an era of advanced AI","volume":"40","author":"Abolghasemi","year":"2024","journal-title":"International Journal of Forecasting"},{"issue":"1","key":"2026102102081514400_ref002","first-page":"1","article-title":"Random forest for tourism demand forecasting: a comparative study","volume":"95","author":"Anshori","year":"2025","journal-title":"Journal of Statistical Computation and Simulation"},{"issue":"3","key":"2026102102081514400_ref003","first-page":"1249","article-title":"The tourism forecasting competition","volume":"39","author":"Athanasopoulos","year":"2023","journal-title":"International Journal of Forecasting"},{"issue":"2","key":"2026102102081514400_ref004","first-page":"456","article-title":"Extrapolation limitations of tree-based methods in tourism forecasting","volume":"30","author":"Bouhaddour","year":"2024","journal-title":"Tourism Economics"},{"key":"2026102102081514400_ref005","doi-asserted-by":"crossref","first-page":"3","DOI":"10.1016\/j.tourman.2017.05.009","article-title":"Structural breaks in international tourism demand: are they caused by crises or disasters?","volume":"63","author":"Cr\u00f3","year":"2017","journal-title":"Tourism Management"},{"issue":"1","key":"2026102102081514400_ref006","doi-asserted-by":"crossref","first-page":"29","DOI":"10.1016\/0261-5177(94)00005-U","article-title":"An integrative approach to tourism demand forecasting","volume":"16","author":"Faulkner","year":"1995","journal-title":"Tourism Management"},{"key":"2026102102081514400_ref007","article-title":"Large language models are zero-shot time series forecasters","volume":"36","author":"Gruver","year":"2024","journal-title":"Advances in Neural Information Processing Systems"},{"issue":"2","key":"2026102102081514400_ref008","doi-asserted-by":"crossref","first-page":"62","DOI":"10.3390\/bdcc7020062","article-title":"The role of ChatGPT in data science: how AI-assisted conversational interfaces are revolutionizing the field","volume":"7","author":"Hassani","year":"2024","journal-title":"Big Data and Cognitive Computing"},{"key":"2026102102081514400_ref009","unstructured":"Jin, M., Wang, S., Ma, L., Chu, Z., Zhang, J.Y., Shi, X., Chen, P.-Y., Liang, Y., Li, Y.-F., Pan, S. and Wen, Q. (2024), \u201cTime-LLM: time series forecasting by reprogramming large language models\u201d, arXiv preprintarXiv:2310.01728."},{"issue":"3","key":"2026102102081514400_ref010","doi-asserted-by":"crossref","first-page":"325","DOI":"10.1080\/19388160.2013.812901","article-title":"Improving forecasting accuracy by combining statistical and judgmental forecasts in tourism","volume":"9","author":"Lin","year":"2013","journal-title":"Journal of China Tourism Research"},{"key":"2026102102081514400_ref011","first-page":"208","article-title":"Accuracy and bias of experts\u2019 adjusted forecasts","volume":"44","author":"Lin","year":"2014","journal-title":"Annals of Tourism Research"},{"key":"2026102102081514400_ref012","volume-title":"Forecasting: Methods and Applications","author":"Makridakis","year":"2008","edition":"3rd ed."},{"key":"2026102102081514400_ref013","doi-asserted-by":"crossref","first-page":"181","DOI":"10.1016\/j.tourman.2014.04.005","article-title":"A meta-analysis of international tourism demand forecasting and implications for practice","volume":"45","author":"Peng","year":"2021","journal-title":"Tourism Management"},{"issue":"1","key":"2026102102081514400_ref014","first-page":"45","article-title":"Exponential smoothing methods for tourism forecasting: a comparative analysis","volume":"10","author":"Sinani","year":"2024","journal-title":"Journal of Tourism Futures"},{"issue":"2","key":"2026102102081514400_ref015","doi-asserted-by":"crossref","first-page":"203","DOI":"10.1016\/j.tourman.2007.07.016","article-title":"Tourism demand modelling and forecasting: a review of recent research","volume":"29","author":"Song","year":"2008","journal-title":"Tourism Management"},{"key":"2026102102081514400_ref016","first-page":"89","volume-title":"International Handbook on the Economics of Tourism","author":"Song","year":"2006"},{"issue":"2","key":"2026102102081514400_ref017","doi-asserted-by":"crossref","first-page":"295","DOI":"10.1016\/j.ijforecast.2011.12.003","article-title":"Combining statistical and judgmental forecasts via a web-based tourism demand forecasting system","volume":"29","author":"Song","year":"2013","journal-title":"International Journal of Forecasting"},{"issue":"1","key":"2026102102081514400_ref018","doi-asserted-by":"crossref","first-page":"37","DOI":"10.1080\/00031305.2017.1380080","article-title":"Forecasting at scale","volume":"72","author":"Taylor","year":"2018","journal-title":"The American Statistician"},{"issue":"3","key":"2026102102081514400_ref019","doi-asserted-by":"crossref","first-page":"447","DOI":"10.1016\/0169-2070(95)00591-7","article-title":"Forecasting tourism demand: a review of empirical research","volume":"11","author":"Witt","year":"1995","journal-title":"International Journal of Forecasting"},{"key":"2026102102081514400_ref020","first-page":"104728","article-title":"ChatGPT and tourism demand forecasting: evidence from Macau","volume":"97","author":"Wu","year":"2025","journal-title":"Tourism Management"},{"key":"2026102102081514400_ref021","doi-asserted-by":"crossref","first-page":"103149","DOI":"10.1016\/j.annals.2021.103149","article-title":"Forecasting tourism recovery amid COVID-19","volume":"87","author":"Zhang","year":"2021","journal-title":"Annals of Tourism Research"},{"issue":"10","key":"2026102102081514400_ref022","doi-asserted-by":"crossref","first-page":"7585","DOI":"10.1016\/j.aej.2022.01.011","article-title":"Comparative analysis of gated recurrent units (GRU), long Short-Term memory (LSTM) cells, autoregressive integrated moving average (ARIMA), seasonal autoregressive integrated moving average (SARIMA) for forecasting COVID-19 trends","volume":"61","author":"ArunKumar","year":"2022","journal-title":"Alexandria Engineering Journal"},{"issue":"1","key":"2026102102081514400_ref023","doi-asserted-by":"crossref","first-page":"209","DOI":"10.5367\/te.2011.0030","article-title":"Research note: tourism demand forecasting with SARIMA models \u2013 the case of South Tyrol","volume":"17","author":"Brida","year":"2011","journal-title":"Tourism Economics"},{"issue":"1","key":"2026102102081514400_ref024","first-page":"1","article-title":"Extrapolated random trees for regression","volume":"24","author":"Cai","year":"2023","journal-title":"Journal of Machine Learning Research"},{"issue":"7","key":"2026102102081514400_ref025","first-page":"983","article-title":"Global tourism forecasting models: a systematic review","volume":"25","author":"Chen","year":"2019","journal-title":"Tourism Economics"},{"key":"2026102102081514400_ref026","first-page":"105899","article-title":"Limitations of tree-based methods for extrapolation in environmental modeling","volume":"171","author":"Hateffard","year":"2024","journal-title":"Environmental Modelling and Software"},{"key":"2026102102081514400_ref027","doi-asserted-by":"crossref","first-page":"25","DOI":"10.1016\/j.jhtm.2021.08.022","article-title":"Using SARIMA-CNN-LSTM approach to forecast daily tourism demand","volume":"49","author":"He","year":"2021","journal-title":"Journal of Hospitality and Tourism Management"},{"issue":"1","key":"2026102102081514400_ref028","doi-asserted-by":"crossref","first-page":"5","DOI":"10.1016\/j.ijforecast.2003.09.015","article-title":"Forecasting seasonals and trends by exponentially weighted moving averages","volume":"20","author":"Holt","year":"2004","journal-title":"International Journal of Forecasting"},{"issue":"3","key":"2026102102081514400_ref029","doi-asserted-by":"crossref","first-page":"439","DOI":"10.1016\/S0169-2070(01)00110-8","article-title":"A state space framework for automatic forecasting using exponential smoothing methods","volume":"18","author":"Hyndman","year":"2002","journal-title":"International Journal of Forecasting"},{"issue":"3","key":"2026102102081514400_ref030","doi-asserted-by":"crossref","first-page":"301","DOI":"10.5367\/000000005774353006","article-title":"Neural network forecasting of tourism demand","volume":"11","author":"Kon","year":"2005","journal-title":"Tourism Economics"},{"issue":"3","key":"2026102102081514400_ref031","doi-asserted-by":"crossref","first-page":"233","DOI":"10.1016\/0261-5177(87)90055-0","article-title":"Tourism demand forecasting models: choice of appropriate variable to represent tourists\u2019 cost of living","volume":"8","author":"Martin","year":"1987","journal-title":"Tourism Management"},{"issue":"2","key":"2026102102081514400_ref032","first-page":"145","article-title":"Limitations of SARIMA models under structural breaks in tourism demand","volume":"29","author":"Montesinos","year":"2022","journal-title":"Journal of Tourism Analysis"},{"issue":"5","key":"2026102102081514400_ref033","first-page":"1335","article-title":"Machine learning approaches for tourism demand forecasting: a review","volume":"28","author":"\u00d6nder","year":"2022","journal-title":"Tourism Economics"},{"issue":"1","key":"2026102102081514400_ref034","doi-asserted-by":"crossref","first-page":"63","DOI":"10.5367\/000000010790872213","article-title":"Tourism demand modelling and forecasting: how should demand be measured?","volume":"16","author":"Song","year":"2019","journal-title":"Tourism Economics"},{"issue":"5","key":"2026102102081514400_ref035","doi-asserted-by":"crossref","first-page":"1011","DOI":"10.1177\/1354816616656418","article-title":"Patterns of seasonality and tourism demand forecasting","volume":"23","author":"Vergori","year":"2016","journal-title":"Tourism Economics"},{"issue":"3","key":"2026102102081514400_ref036","doi-asserted-by":"crossref","first-page":"324","DOI":"10.1287\/mnsc.6.3.324","article-title":"Forecasting sales by exponentially weighted moving averages","volume":"6","author":"Winters","year":"1960","journal-title":"Management Science"},{"key":"2026102102081514400_ref037","first-page":"205","article-title":"Limitations of random forests for extrapolation in regression tasks","volume":"92","author":"Zhang","year":"2019","journal-title":"Pattern Recognition"}],"container-title":["Tourism Review"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.emerald.com\/tr\/article-pdf\/doi\/10.1108\/TR-02-2026-0223\/11921307\/tr-02-2026-0223en.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/www.emerald.com\/tr\/article-pdf\/doi\/10.1108\/TR-02-2026-0223\/11921307\/tr-02-2026-0223en.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,10,21]],"date-time":"2026-10-21T06:08:28Z","timestamp":1792562908000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.emerald.com\/tr\/article\/doi\/10.1108\/TR-02-2026-0223\/1403534\/Tourism-demand-forecasting-in-the-artificial"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,10,22]]},"references-count":37,"URL":"https:\/\/doi.org\/10.1108\/tr-02-2026-0223","relation":{},"ISSN":["1660-5373","1759-8451"],"issn-type":[{"value":"1660-5373","type":"print"},{"value":"1759-8451","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,10,22]]}}}