Analysis of mobility as a service solution option in tourism traffic scenarios
SUN Yilin1,2, CHEN Shujie1, HUANG Pei2
1. Polytechnic Institute, Zhejiang University, Hangzhou 310058, China; 2. College of Civil Engineering and Architecture, Zhejiang University, Hangzhou 310058, China
摘要出行即服务(mobility as a service,MaaS)是基于通信技术的城市交通网络系统的重要组成部分,为推进城市交通向可持续发展转型提供综合出行信息支持。该文基于旅游出行网络问卷调查数据,构建嵌套Logit模型,研究旅行者出行方式及MaaS方案选择行为及其影响因素。为进一步揭示影响因素与MaaS方案选择间的非线性关系,建立轻量级梯度提升机(light gradient boosting machine,LightGBM),并结合SHAP (shapley additive explanations)机器学习模型解释方法,剖析主要影响因素之间的协同效应及其对MaaS方案选择的作用。结果表明,收入、职业、家庭生命周期、平日交通出行习惯、景点类型等显著影响旅行者对MaaS方案的选择,其中低收入群体、游览商业文娱和文化遗迹景点的旅行者更偏好含不限次公交或地铁的MaaS方案等,因此可以为特定群体个性化制定相关优惠措施,丰富MaaS在旅游交通中的相关实践。研究结论对于促进公共交通的使用、提升多式联运城市交通网络的运行效率及旅游交通吸引力具有借鉴意义。
Abstract：[Objective] Mobility as a service (MaaS) plays a critical role in urban transportation networks, leveraging communication technologies to provide integrated travel information while promoting sustainable development. Despite its importance, the application of MaaS in tourism traffic still needs to be explored.[Methods] This study focused on tourists visiting Hangzhou and designed a survey questionnaire for data collection. A combined approach of stated preference and revealed preference methods was employed to examine travel behavior within the context of tourism traffic. Using data from an online platform, the study analyzed tourists' travel mode choices and their preferences for different MaaS solutions. First, a nested Logit model was developed, with the upper level representing multimodal transportation choices while the lower level analyzed their MaaS solution selections. Influencing factors were categorized into three main groups: individual characteristics, household attributes, and travel-related factors. Given the potential nonlinear relationships between these influencing factors and MaaS solution choices, a light gradient boosting machine model was further applied. This machine learning model explored interactive effects between various factors on MaaS choice, such as how age and income jointly influence MaaS preferences or how the choice of tourist attractions relates to tourist demographics. To further elucidate the collaborative effects of key factors, the Shapley additive explanations method was incorporated to interpret the collaborative effects of multiple factors on MaaS decisions.[Results] The results indicate a strong connection between MaaS solution choices and tourists' preferred travel modes, validating the use of the nested logit model. Several key factors emerged as significant influencers of MaaS preferences, including income, occupation, household lifecycle, daily travel habits, and the type of tourist attractions visited. Specifically, self-employed individuals, those who frequently use public transport, and tourists who plan fewer bus trips showed a preference for MaaS solutions combining discounts on subways and taxis. Groups traveling with elderly individuals or those accustomed to “public transport and private car” habits leaned toward MaaS solutions that offer unlimited subway rides. Additionally, the type of tourist attractions plays a crucial role in shaping MaaS preferences. Tourists visiting commercial, entertainment, or cultural heritage sites, particularly from lower-income groups, chose MaaS solutions with unlimited bus rides. Meanwhile, those visiting historical landmarks favored taxi-centered MaaS options. Tourists aged 45 and above were less likely to select MaaS solutions with unlimited subway rides as their income rose. Older tourists visiting historical sites showed a stronger preference for MaaS solutions, prioritizing taxi services. Public transport users who gravitate toward subway-based MaaS solutions share common traits, such as being in low-income groups, self-employed, traveling with elderly companions, and regularly transferring between public transport modes in their daily routines.[Conclusions] These findings can guide public transportation agencies, tourism operators, and other stakeholders in designing user-centered MaaS solutions. By segmenting users based on demographics and travel habits, personalized travel services can be created to improve the applicability of MaaS in tourism transportation. The conclusions of this study have significant implications for promoting the use of public transportation, improving the operational efficiency of multimodal urban transportation networks, and increasing the attractiveness of transportation options for tourists.
孙轶琳, 谌淑杰, 黄佩. 旅游交通场景下出行即服务方案选择分析[J]. 清华大学学报（自然科学版）, 2025, 65(5): 959-969.
SUN Yilin, CHEN Shujie, HUANG Pei. Analysis of mobility as a service solution option in tourism traffic scenarios. Journal of Tsinghua University(Science and Technology), 2025, 65(5): 959-969.
[1] 中华人民共和国文化和旅游部. 2023年国内旅游数据情况[EB/OL].(2024-02-10)[2024-08-10] . https://www.gov.cn/lianbo/bumen/202402/content_6931178.htm. Ministry of Culture and Tourism of the People's Republic of China. Domestic tourism data situation in 2023[EB/OL].(2024-02-10)[2024-08-10] . https://www.gov.cn/lianbo/bumen/202402/content_6931178.htm.(in Chinese) [2] POLYDOROPOULOU A, PAGONI I, TSIRIMPA A, et al. Prototype business models for Mobility-as-a-Service[J]. Transportation Research Part A:Policy and Practice, 2020, 131:149-162. [3] 王月,姚恩建,郝赫.低碳导向的多模式交通出行服务定价策略[J].清华大学学报(自然科学版), 2023, 63(11):1741-1749. WANG Y, YAO E J, HAO H. Low-carbon-oriented pricing strategy of multi-mode transportation service[J]. Journal of Tsinghua University (Science and Technology), 2023, 63(11):1741-1749.(in Chinese) [4] 李婉莹,关宏志,韩艳,等.社会偏好视野下旅游出行服务链定价博弈模型[J].交通运输系统工程与信息, 2022, 22(4):11-22. LI W Y, GUAN H Z, HAN Y, et al. Pricing game model of travel service chain under social preference view[J]. Journal of Transportation Systems Engineering and Information Technology, 2022, 22(4):11-22.(in Chinese) [5] MATYAS M, KAMARGIANNI M. The potential of mobility as a service bundles as a mobility management tool[J]. Transportation, 2019, 46(5):1951-1968. [6] CAIATI V, RASOULI S, TIMMERMANS H. Bundling, pricing schemes and extra features preferences for mobility as a service:Sequential portfolio choice experiment[J]. Transportation Research Part A:Policy and Practice, 2020, 131:123-148. [7] HO C Q, MULLEY C, HENSHER D A. Public preferences for mobility as a service:Insights from stated preference surveys[J]. Transportation Research Part A:Policy and Practice, 2020, 131:70-90. [8] KIM Y, KIM E J, JANG S, et al. A comparative analysis of the users of private cars and public transportation for intermodal options under Mobility-as-a-Service in Seoul[J]. Travel Behaviour and Society, 2021, 24:68-80. [9] 杨敏,李宏伟,任怡凤,等.基于旅客异质性画像的公铁联程出行方案推荐方法[J].清华大学学报(自然科学版), 2022, 62(7):1220-1227. YANG M, LI H W, REN Y F, et al. Road-rail intermodal travel recommendations based on a passenger heterogeneity profile[J]. Journal of Tsinghua University (Science and Technology), 2022, 62(7):1220-1227.(in Chinese) [10] VIJ A, RYAN S, SAMPSON S, et al. Consumer preferences for Mobility-as-a-Service (MaaS) in Australia[J]. Transportation Research Part C:Emerging Technologies, 2020, 117:102699. [11] LILJAMO T, LIIMATAINEN H, PÖLLÄNEN M, et al. People's current mobility costs and willingness to pay for Mobility as a Service offerings[J]. Transportation Research Part A:Policy and Practice, 2020, 136:99-119. [12] GUIDON S, WICKI M, BERNAUER T, et al. Transportation service bundling-For whose benefit?Consumer valuation of pure bundling in the passenger transportation market[J]. Transportation Research Part A:Policy and Practice, 2020, 131:91-106. [13] KRAUSS K, RECK D J, AXHAUSEN K W. How does transport supply and mobility behaviour impact preferences for MaaS bundles?A multi-city approach[J]. Transportation Research Part C:Emerging Technologies, 2023, 147:104013. [14] KIM E J, KIM Y, JANG S, et al. Tourists'preference on the combination of travel modes under Mobility-as-a-Service environment[J]. Transportation Research Part A:Policy and Practice, 2021, 150:236-255. [15] JANG S, CAIATI V, RASOULI S, et al. Does MaaS contribute to sustainable transportation?A mode choice perspective[J]. International Journal of Sustainable Transportation, 2021, 15(5):351-363. [16] RATILAINEN H. Mobility-as-a-Service:Exploring consumer preferences for MaaS subscription packages using a stated choice experiment[D]. Delft:Delft University of Technology, 2017. [17] LE-KLÄHN D T, ROOSEN J, GERIKE R, et al. Factors affecting tourists'public transport use and areas visited at destinations[J]. Tourism Geographies, 2015, 17(5):738-757. [18] 世界资源研究所.出行即服务(MAAS)实践指南介绍与案例集[EB/OL].(2022-11-10)[2024-08-10] . https://wri.org.cn/report/MaaS-Guideline-for-Chinese-Cities-and-Case-Studies. World Resources Institute. MaaS practice guidelines and case studies[EB/OL].(2022-11-10)[2024-08-10] . https://wri.org.cn/report/MaaS-Guideline-for-Chinese-Cities-and-Case-Studies.(in Chinese) [19] HO C Q, HENSHER D A, RECK D J, et al. MaaS bundle design and implementation:Lessons from the Sydney MaaS trial[J]. Transportation Research Part A:Policy and Practice, 2021, 149:339-376. [20] WEN C H, KOPPELMAN F S. The generalized nested logit model[J]. Transportation Research Part B:Methodological, 2001, 35(7):627-641. [21] 关宏志.非集计模型:交通行为分析的工具[M].北京:人民交通出版社, 2004. GUAN H Z. Disaggregate model:A tool of traffic behavior analysis[M]. Beijing:China Communications Press, 2004.(in Chinese) [22] BIERLAIRE M. Calculating indicators with PandasBiogeme[J]. Transport and Mobility Laboratory, Ecole Polytechnique Fédérale de Lausanne:Technical report, 2018. [23] AL DAOUD E. Comparison between XGBoost, LightGBM and CatBoost using a home credit dataset[J]. International Journal of Computer and Information Engineering, 2019, 13(1):6-10. [24] SUN X L, LIU M X, SIMA Z Q. A novel cryptocurrency price trend forecasting model based on LightGBM[J]. Finance Research Letters, 2020, 32:101084. [25] 李跞.基于GAN和LightGBM的个人信用风险预测模型的研究与应用[D].重庆:重庆大学, 2022. LI L. Research and implementation of personal credit risk prediction model based on GAN and LightGBM[D]. Chongqing:Chongqing University, 2022.(in Chinese) [26] LUNDBERG S M, LEE S I. A unified approach to interpreting model predictions[C]//Proceedings of the 31st International Conference on Neural Information Processing Systems. Red Hook, USA:Curran Associates Inc., 2017:4768-4777. [27] LI Y G, YAO E J, YANG Y, et al. Modeling the tourism travel mode and route choice behaviour based on nested logit model[C]//2020 IEEE 5th International Conference on Intelligent Transportation Engineering (ICITE). Beijing, China:IEEE, 2020:28-32.
2025, Vol. 65 
