All Issue

2026 Vol.44, Issue 2 Preview Page

Article

Synthesizing Human Mobility Patterns Using Large Language Models - A Sejong City Case Study

거대 언어 모델을 활용한 통행 생성 방법 개발: 세종시를 중심으로

Chiwoo ROH1
,
Minsu WON2*
노 치우1
,
원 민수2*
1Researcher, Department of Transport Big Data, Korea Transport Institute, Sejong 30147, Korea
2Senior Research Fellow, Department of Transport Big Data, Korea Transport Institute, Sejong 30147, Korea
1한국교통연구원 교통빅데이터연구본부 연구원
2한국교통연구원 교통빅데이터연구본부 연구위원
*Corresponding Author
30 April 2026. pp. 270-288
PDF XML Citation
Abstract

This study investigated whether large language models (LLMs) can generate realistic personal activity schedules without additional training by leveraging their pre-trained knowledge and contextual reasoning capabilities. Previous trip generation research has shifted from trip-based approaches to activity-based models (ABMs), but these models face limitations due to their reliance on high-resolution travel diary data. LLMs capable of reflecting human decision-making contexts based on pre-trained knowledge have emerged, and this study explored their potential for generating personal activity schedules. Using Sejong City's individual travel survey data, the study compared trip generation performance across four LLM utilization strategies: Zero-shot, Few-shot, Few-shot with statistics, and Retrieval-Augmented Generation (RAG). Using Jensen-Shannon Divergence (JSD) to compare and analyze the similarity of probability distributions between reference and generated data, the LLM was found capable of reproducing realistic trip distributions overall without additional retraining. Specifically, the RAG strategy demonstrated the highest spatial and temporal distribution reproduction performance, as well as the best reproduction of rare trip patterns, followed by Few-shot with statistics, zero-shot, and few-shot. These results suggest that LLMs can generate trip patterns without additional training, suggesting it could serve as a new trip generation method applicable even in data-scarce environments.

Keywords
large language models
prompt design
retrieval augmented generation
trip diary
trip generation

본 연구는 거대 언어 모델(Large Language Model, LLM)의 사전 학습지식과 맥락적 추론 능력을 활용하여, 추가 학습 없이도 현실적인 개인 통행 일정을 생성할 수 있는지를 확인하였다. 기존 통행 생성 연구의 경우 Trip-based 방식 접근에서 활동 기반 모형(Activity-Based Model, ABM) 중심으로 발전해왔으나, 고해상도 통행일지 데이터 의존의 한계를 지닌다. 사전학습된 지식을 기반으로 인간의 의사결정 맥락을 반영할 수 있는 LLM이 등장하였고, 이를 활용하여 개인 통행 일정 생성 가능성을 확인하였다. 본 연구는 세종시 개인통행실태조사 데이터를 활용하여, Zero-shot, Few-shot, Few-shot with statistics, Retrieval-Augmented Generation(RAG) 등 네 가지 LLM 사용 전략에 따른 통행 생성 성능을 비교하였다. Jensen-Shannon Divergence(JSD)를 이용해 기준 데이터와 생성 데이터의 확률분포 유사도를 비교‧분석한 결과, LLM은 별도의 재학습 없이도 전반적으로 현실적인 통행 분포를 재현할 수 있었으며, 특히 RAG 전략이 가장 높은 공간·시간 분포 및 희귀 통행 패턴 재현 성능을 보여주었으며 Few-shot with statistics, zero-shot, few-shot 순으로 뒤따랐다. 이러한 결과들을 바탕으로 LLM이 추가 학습 없이도 통행 생성을 수행할 수 있다는 점에서, 데이터 희소 환경에서도 활용가능한 새로운 통행 생성 방법의 가능성이 될 수 있음을 확인하였다.

키워드
거대 언어 모델
프롬프트 디자인
검색 증강 생성
통행 일지
통행 생성
References
1

Achiam J., Adler S., Agarwal S., Ahmad L., Akkaya I., Aleman F. L. et al. (2023), GPT-4 Technical Report, arXiv preprint arXiv:2303.08774.

2

Arentze T., Hofman F., Van Mourik H., Timmermans H. (2000), ALBATROSS: Multiagent, Rule-Based Model of Activity Pattern Decisions, Transportation Research Record, 1706(1), 136-144.

10.3141/1706-16
3

Barbosa H., Barthelemy M., Ghoshal G., James C. R., Lenormand M., Louail T. et al. (2018), Human Mobility: Models and Applications, Physics Reports, 734, 1-74.

10.1016/j.physrep.2018.01.001
4

Ben-Akiva M. E., Bowman J. L. (1998), Activity Based Travel Demand Model Systems, In Equilibrium and Advanced Transportation Modelling, Springer US (Boston, MA), 27-46.

10.1007/978-1-4615-5757-9_2
5

Beneduce C., Lepri B., Luca M. (2025), Large Language Models Are Zero-Shot Next Location Predictors, IEEE Access.

10.1109/ACCESS.2025.3565297
6

Bhat C. R., Koppelman F. S. (1999), Activity-Based Modeling of Travel Demand, In Handbook of Transportation Science, Springer, 35-61.

10.1007/978-1-4615-5203-1_3
7

Bowman J. L. (1995), Activity Based Travel Demand Model System With Daily Activity Schedules, Doctoral Dissertation, Massachusetts Institute of Technology.

8

Bowman J. L., Bradley M. (2017), Testing Spatial Transferability of Activity-Based Travel Forecasting Models, Transportation Research Record, 2669, 62-71.

10.3141/2669-07
9

Chakraborty S., Tomsett R., Raghavendra R., Harborne D., Alzantot M., Cerutti F. et al. (2017), Interpretability of Deep Learning Models: A Survey of Results, In 2017 IEEE SmartWorld, Ubiquitous Intelligence & Computing, Advanced & Trusted Computing, Scalable Computing & Communications, Cloud & Big Data Computing, Internet of People and Smart City Innovation, IEEE, 1-6.

10.1109/UIC-ATC.2017.8397411
10

Gao Y., Xiong Y., Gao X., Jia K., Pan J., Bi Y. et al. (2023), Retrieval-Augmented Generation for Large Language Models: A Survey, arXiv preprint arXiv:2312.10997.

11

Gonzales C. A., Salazar D., Niño Uy S., Lim R. (2025), Hybrid Deep Learning and Agent-Based Modeling for Dynamic Urban Traffic Forecasting in Smart Cities, 4(1), 48-62.

10.63876/ijtm.v4i1.129
12

Goulias K. G., Bhat C. R., Pendyala R. M., Chen Y., Paleti R., Konduri K. C. et al. (2011), Simulator of Activities, Greenhouse Emissions, Networks, and Travel (SimAGENT) in Southern California: Design, Implementation, Preliminary Findings, and Integration Plans, In 2011 IEEE Forum on Integrated and Sustainable Transportation Systems, IEEE, 164-169.

10.1109/FISTS.2011.5973624
13

Hassan M., Kabir M. E., Jusoh M., An H. K., Negnevitsky M., Li C. (2025), Large Language Models in Transportation: A Comprehensive Bibliometric Analysis of Emerging Trends, Challenges and Future Research, IEEE Access.

10.1109/ACCESS.2025.3589319
14

Huang Z., Ling X., Wang P., Zhang F., Mao Y., Lin T. et al. (2018), Modeling Real-Time Human Mobility Based on Mobile Phone and Transportation Data Fusion, Transportation Research Part C: Emerging Technologies, 96, 251-269.

10.1016/j.trc.2018.09.016
15

Joubert J. W., De Waal A. (2020), Activity-Based Travel Demand Generation Using Bayesian Networks, Transportation Research Part C: Emerging Technologies, 120, 102804.

10.1016/j.trc.2020.102804
16

Korea Transport Database (국가교통DB), https://www.ktdb.go.kr/www/contents.do?key=15, 2025.11.11.

17

Kulkarni A., McNally M. G. (2000), An Activity-Based Travel Pattern Generation Model, Institute of Transportation Studies, University of California (Irvine, USA).

18

Kullback S., Leibler R. A. (1951), On Information and Sufficiency, The Annals of Mathematical Statistics, 22(1), 79-86.

10.1214/aoms/1177729694
19

Lee G., Choi K. (2006), Estimation of Induced Highway Travel Demand, J. Korean Soc. Transp., 24(7), Korean Society of Transportation, 91-100.

20

Lewis P., Perez E., Piktus A., Petroni F., Karpukhin V., Goyal N. et al. (2020), Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks, Advances in Neural Information Processing Systems, 33, 9459-9474.

21

Li S., Feng J., Chi J., Hu X., Zhao X., Xu F. (2024a), LIMP: Large Language Model Enhanced Intent-Aware Mobility Prediction, arXiv preprint arXiv:2408.12832.

22

Li X., Huang F., Lv J., Xiao Z., Li G., Yue Y. (2024b), Be More Real: Travel Diary Generation Using LLM Agents and Individual Profiles, arXiv preprint arXiv:2407.18932.

23

Li Y., Li M., Rice M., Zhang H., Sha D., Li M. et al. (2021), The Impact of Policy Measures on Human Mobility, COVID-19 Cases, and Mortality in the US: A Spatiotemporal Perspective, International Journal of Environmental Research and Public Health, 18(3), 996.

10.3390/ijerph1803099633498647PMC7908236
24

Liang Y., Liu Y., Wang X., Zhao Z. (2024), Exploring Large Language Models for Human Mobility Prediction Under Public Events, Computers, Environment and Urban Systems, 112, 102153.

10.1016/j.compenvurbsys.2024.102153
25

Lim K., Kim S., Chung S. (2013), Activity-Based Approaches for Travel Demand Modeling: Reviews on Developments and Implementations, KSCE Journal of Civil and Environmental Engineering Research, 33(2), 719-727.

10.12652/Ksce.2013.33.2.719
26

Lin J. (1991), Divergence Measures Based on the Shannon Entropy, IEEE Transactions on Information Theory, 37(1), 145-151.

10.1109/18.61115
27

Liu T., Yang J., Yin Y. (2025), Toward LLM-Agent-Based Modeling of Transportation Systems: A Conceptual Framework, Artificial Intelligence for Transportation, 1, 100001.

10.1016/j.ait.2025.100001
28

Ma H., Liu Y., Jiang Q., He B. Y., Liao X., Ma J. (2024), Mobility AI Agents and Networks, IEEE Transactions on Intelligent Vehicles, 9(7), 5124-5129.

10.1109/TIV.2024.3454285
29

Masri S., Ashqar H. I., Elhenawy M. (2024), Leveraging Large Language Models (LLMs) for Traffic Management at Urban Intersections: The Case of Mixed Traffic Scenarios, arXiv preprint arXiv:2408.00948.

10.1109/ICUFN65838.2025.11170100
30

McNally, M. G. (1996). An activity-based microsimulation model for travel demand forecasting.

31

Molla M. M., Stone M. L., Motuba D. (2017), Developing an Activity-Based Trip Generation Model for Small/Medium Size Planning Agencies, Transportation Planning and Technology, 40(5), 540-555.

10.1080/03081060.2017.1314505
32

OpenAI, https://openai.com/index/introducing-gpt-5/, 2025.11.11.

33

Pellungrini R., Pappalardo L., Pratesi F., Monreale A. (2017), Fast Estimation of Privacy Risk in Human Mobility Data, In International Conference on Computer Safety, Reliability, and Security, Springer International Publishing (Cham, Switzerland), 415-426.

10.1007/978-3-319-66284-8_35
34

Prelipcean A. C., Susilo Y. O., Gidofalvi G. (2018), Future Directions of Research for Automatic Travel Diary Collection, In 11th International Conference on Transport Survey Methods.

35

Tang J., Liu F., Wang Y., Wang H. (2015), Uncovering Urban Human Mobility From Large Scale Taxi GPS Data, Physica A: Statistical Mechanics and its Applications, 438, 140-153.

10.1016/j.physa.2015.06.032
36

Wang J., Jiang R., Yang C., Wu Z., Onizuka M., Shibasaki R. et al. (2024), Large Language Models as Urban Residents: An LLM Agent Framework for Personal Mobility Generation, Advances in Neural Information Processing Systems, 37, 124547-124574.

10.52202/079017-3957
37

Wang X., Fang M., Zeng Z., Cheng T. (2023), Where Would I Go Next? Large Language Models as Human Mobility Predictors, arXiv preprint arXiv:2308.15197.

38

Ying S., Li Z., Yu M. (2025), Beyond Words: Evaluating Large Language Models in Transportation Planning, Geo-spatial Information Science, 1-23.

10.1080/10095020.2025.2493073
39

Zhang L., Levinson D. (2004), Agent-Based Approach to Travel Demand Modeling: Exploratory Analysis, Transportation Research Record, 1898, 28-36.

10.3141/1898-04
40

Zhang L., Yang D., Ghader S., Carrion C., Xiong C., Rossi T. F. et al. (2018), An Integrated, Validated, and Applied Activity-Based Dynamic Traffic Assignment Model for the Baltimore-Washington Region, Transportation Research Record, 2672(51), 45-55.

10.1177/0361198118796397
41

Zhao B., Kumar K., Casey G., Soga K. (2019), Agent-Based Model (ABM) for City-Scale Traffic Simulation: A Case Study on San Francisco, In International Conference on Smart Infrastructure and Construction 2019, ICE Publishing, 203-212.

10.1680/icsic.64669.203
Information
  • Publisher :Korean Society of Transportation
  • Publisher(Ko) :대한교통학회
  • Journal Title :Journal of Korean Society of Transportation
  • Journal Title(Ko) :대한교통학회지
  • Volume : 44
  • No :2
  • Pages :270-288
  • Received Date : 2025-12-05
  • Revised Date : 2025-12-15
  • Accepted Date : 2026-01-20
  • DOI :https://doi.org/10.7470/jkst.2026.44.2.270
Journal Informaiton Journal of Korean Society of Transportation Journal of Korean Society of Transportation
Archives
: 2015 Vol.34~
  • NRF
  • KOFST
  • crossref cited-by
  • crosscheck
  • orcid
  • open access
  • ccl
Journal Informaiton Journal Informaiton - close