Optimization of Tourism Consumption Prediction Model Integrating LSTM-Attention
Main Article Content
Abstract
This paper addresses seasonal structural drift and window-selection limitations in tourism consumption forecasting by proposing a BiLSTM-Att+PE model to capture dynamic seasonal patterns and improve consumption-pattern representation. The model combines BiLSTM for bidirectional temporal dependency learning, Bahdanau attention for dynamic focus on key time steps such as holidays and emergencies, and learnable positional encoding for seasonal information modeling. It is trained end-to-end on provincial tourism economic data from 2001Q1 to 2024Q4, including consumption, visitor numbers, GDP, income, holidays, and epidemic markers. Four sample sets are constructed through sliding windows with T ∈{1, 2, 3, 4} quarters. The model is optimized using MSE loss and AdamW, while Bayesian search determines the best hyperparameters, including a learning rate of 3.4 × 10−4, 128 LSTM units, and 64 attention dimensions. The optimal performance is achieved with a two-quarter window, reaching RMSE of 21.3 billion and R2 of 0.962, outperforming seven benchmarks. Attention visualization confirms adaptive focus on holiday and epidemic events, providing an interpretable and computationally efficient framework for nonlinear time-series forecasting.
Downloads
Article Details
This work is licensed under a Creative Commons Attribution 4.0 International License.
Authors who publish with this journal agree to the following terms:
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See The Effect of Open Access).
References
J. Sui and Y. Zhang, “Identification of dynamic path evolution of regional tourism economic cycle in China,” Tourism Tribune/Lvyou Xuekan, vol. 35, no. 1, pp. 63, 2020, doi: 10.19765/j.cnki.1002-5006.2020.01.010.
D. Xing and G. Sun, “Interactive relationship and differences between residents’ consumption upgrade and tourism consumption structure among provinces in China,” Economic Geography, vol. 44, no. 3, pp. 197-208, 2024, doi: 10.15957/j.cnki.jjdl.2024.03.020.
Q. Xie, M. Lu, and C. Xie, “High-frequency prediction of tourist flow in tourist attractions based on LBS and deep learning,” Journal of Geo-Information Science, vol. 25, no. 2, pp. 298-310, 2023, doi: 10.12082/dqxxkx.2023.220231.
S. C. Hsieh, “Tourism demand forecasting based on an LSTM network and its variants,” Algorithms, vol. 14, no. 8, pp. 243-243, 2021, doi: 10.3390/a14080243.
M. Adil, J. Z. Wu, R. K. Chakrabortty, A. Alahmadi, M. F. Ansari, M. J. Ryan, et al., “Attention-based STL-BiLSTM network to forecast tourist arrival,” Processes, vol. 9, no. 10, pp. 1759-1759, 2021, doi: 10.3390/pr9101759.
L. Wang, “Research on tourism demand prediction algorithm,” Intelligent Computers and Applications, vol. 10, no. 5, pp. 118-118, 2020.
M. Hu, H. Yang, D. C. Wu, and S. Ma, “A novel two-stage combination model for tourism demand forecasting,” Tourism Economics, vol. 30, no. 8, pp. 1925-1950, 2024, doi: 10.1177/13548166241237845.
J. Zhang, Y. Chen, K. Panchamy, G. Jin, C. Wang, L. Yang, et al., “A spatiotemporal integrated prediction model for multi-step short-term passenger flow in urban rail transit network based on attention mechanism,” Journal of Geo-Information Science, vol. 25, no. 4, pp. 698-713, 2023.
X. Li, X. Zhang, L. Yu, and S. Wang, “Research on short-term passenger flow prediction in scenic spots based on improved Transformer model,” Chinese Journal of Management Science, vol. 33, no. 2, pp. 105-117, 2025, doi: 10.16381/j.cnki.issn1003-207x.2023.1927.
X. Meng and H. Shi, “A review of time series data prediction methods based on Transformer model,” Journal of Frontiers of Computer Science & Technology, vol. 19, no. 1, pp. 45-45, 2025, doi: 10.3778/j.issn.1673-9418.2403070.
Q. Feng, X. Sun, and J. Hao, “Dynamic integrated time series prediction method based on state transition regression,” Chinese Journal of Management Science, vol. 32, no. 2, pp. 307-314, 2024, doi: 10.16381/j.cnki.issn1003-207x.2022.0599.
G. Xing, S. Sun, D. Bi, J. E. Guo, and S. Wang, “Seasonal and trend forecasting of tourist arrivals: An adaptive multiscale ensemble learning approach,” International Journal of Tourism Research, vol. 24, no. 3, pp. 425-442, 2022, doi: 10.1002/jtr.2512.
Q. Yu and X. Liu, “The impact and mechanism of cultural tourism consumption promotion policies under the background of cultural tourism integration: empirical evidence from cultural and tourism consumption pilot demonstration cities,” Journal of East China Normal University (Philosophy and Social Sciences), vol. 57, no. 4, pp. 162-162, 2025, doi: 10.16382/j.cnki.1000-5579.2025.04.013.
Z. Xiong, B. Hu, and K. Zhang, “Digital economy empowers tourism industry development: from the perspective of tourism investment and consumption,” Tourism Tribune/Lvyou Xuekan, vol. 39, no. 12, pp. 29-29, 2024, doi: 10.19765/j.cnki.1002-5006.2024.12.008.
D. R. Toubes, N. Araujo Vila, and J. A. Fraiz Brea, “Changes in consumption patterns and tourist promotion after the COVID-19 pandemic,” Journal of theoretical and applied electronic commerce research, vol. 16, no. 5, pp. 1332-1352, 2021, doi: 10.3390/jtaer16050075.
D. C. Wu, C. Cao, W. Liu, and J. L. Chen, “Impact of domestic tourism on economy under COVID-19: The perspective of tourism satellite accounts,” Annals of Tourism Research Empirical Insights, vol. 3, no. 2, pp. 100055-100055, 2022, doi: 10.1016/j.annale.2022.100055.
S. Polyzos, A. Samitas, and A. E. Spyridou, “Tourism demand and the COVID-19 pandemic: An LSTM approach,” Tourism Recreation Research, vol. 46, no. 2, pp. 175-187, 2021, doi: 10.1080/02508281.2020.1777053.
J. Li, H. Yuan, X. Yu, and T. Hu, “The intelligent evaluation in ice and snow tourism based on LSTM network,” Scientific Reports, vol. 14, no. 1, pp. 17342-17342, 2024, doi: 10.1038/s41598-024-68457-w.
M. Luo, D. Chen, X. Hou, K. Luo, and T. Wang, “Combined CNN-BiLSTM-Att tourism flow prediction based on VMD-MWPE decomposition reconstruction,” Scientific Reports, vol. 15, no. 1, pp. 18003-18003, 2025, doi: 10.1038/s41598-025-02348-6.
W. Han, Y. Li, Y. Li, and T. Huang, “A deep learning model based on multi-source data for daily tourist volume forecasting,” Current Issues in Tourism, vol. 27, no. 5, pp. 768-786, 2024, doi: 10.1080/13683500.2023.2183818.
D. Zhu, R. Jing, Q. Guo, D. Zhang, and F. Wan, “Sentiment analysis of tourism review text combined with bertbilstm and attention mechanism,” Journal of Computational Methods in Sciences and Engineering, vol. 24, no. 3, pp. 1605-1615, 2024, doi: 10.3233/JCM-247135.
Y. Zhang, W. H. Tan, and Z. Zeng, “Tourism Demand Forecasting Based on a Hybrid Temporal Neural Network Model for Sustainable Tourism,” Sustainability, vol. 17, no. 5, pp. 2210-2210, 2025, doi: 10.3390/su17052210.
P. Zhou, X. Xie, Z. Lin, and S. Yan, “Towards understanding convergence and generalization of AdamW,” IEEE transactions on pattern analysis and machine intelligence, vol. 46, no. 9, pp. 6486-6493, 2024, doi: 10.1109/TPAMI.2024.3382294.
L. Yang, “Theoretical Analysis of Adam Optimizer in the Presence of Gradient Skewness,” International Journal of Applied Science, vol. 7, no. 2, pp. 27-27, 2024, doi: 10.30560/ijas.v7n2p27.