LingoTrip: Spatiotemporal context prompt driven large language model for individual trip prediction
Large language models (LLMs) showed superior performance in many language-related tasks. It is promising to model the individual mobility prediction problem as a language model and use pretrained LLMs to predict the individual next trip information (e.g., time and location) for personalized travel r...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-01-01
|
| Series: | Journal of Public Transportation |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S1077291X25000025 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832573259751620608 |
|---|---|
| author | Zhenlin Qin Pengfei Zhang Leizhen Wang Zhenliang Ma |
| author_facet | Zhenlin Qin Pengfei Zhang Leizhen Wang Zhenliang Ma |
| author_sort | Zhenlin Qin |
| collection | DOAJ |
| description | Large language models (LLMs) showed superior performance in many language-related tasks. It is promising to model the individual mobility prediction problem as a language model and use pretrained LLMs to predict the individual next trip information (e.g., time and location) for personalized travel recommendations. Theoretically, it is expected to overcome the common limitations of data-driven prediction models in zero/few shot learning, generalization, and interpretability. The paper proposes a LingoTrip model for predicting individual next trip location by designing the spatiotemporal context prompts for LLMs. The designed prompting strategies enable LLMs to capture implicit land use information (trip purposes), spatiotemporal mobility patterns (choice preferences), and geographical dependencies of the stations used (choice variability). The lingoTrip is validated using Hong Kong Mass Transit Railway trip data by comparing it with the state-of-the-art data-driven mobility prediction models under different training data sizes. Sensitivity analyses are performed for model hyperparameters and their tuning methods to adapt for other datasets. The results show that LingoTrip outperforms data-driven models in terms of prediction accuracy, transferability (between individuals), zero/few shot learning (limited training sample size) and interpretability of predictions. The LingoTrip model can facilitate the effective provision of personalized information for system crowding and disruption contexts (i.e., proactively providing information to targeted individuals). |
| format | Article |
| id | doaj-art-ff65ee1f79594039af452ea2d02c13e1 |
| institution | Kabale University |
| issn | 2375-0901 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | Elsevier |
| record_format | Article |
| series | Journal of Public Transportation |
| spelling | doaj-art-ff65ee1f79594039af452ea2d02c13e12025-02-02T05:26:52ZengElsevierJournal of Public Transportation2375-09012025-01-0127100117LingoTrip: Spatiotemporal context prompt driven large language model for individual trip predictionZhenlin Qin0Pengfei Zhang1Leizhen Wang2Zhenliang Ma3Department of Civil and Architectural Engineering, KTH Royal Institute of Technology, Stockholm 10044, SwedenInstitute of Physics, Henan Academy of Sciences, Zhengzhou 450000, ChinaDepartment of Data Science and Artificial Intelligence, Monash University, Melbourne, Victoria 3800, AustraliaDepartment of Civil and Architectural Engineering, KTH Royal Institute of Technology, Stockholm 10044, Sweden; Corresponding author.Large language models (LLMs) showed superior performance in many language-related tasks. It is promising to model the individual mobility prediction problem as a language model and use pretrained LLMs to predict the individual next trip information (e.g., time and location) for personalized travel recommendations. Theoretically, it is expected to overcome the common limitations of data-driven prediction models in zero/few shot learning, generalization, and interpretability. The paper proposes a LingoTrip model for predicting individual next trip location by designing the spatiotemporal context prompts for LLMs. The designed prompting strategies enable LLMs to capture implicit land use information (trip purposes), spatiotemporal mobility patterns (choice preferences), and geographical dependencies of the stations used (choice variability). The lingoTrip is validated using Hong Kong Mass Transit Railway trip data by comparing it with the state-of-the-art data-driven mobility prediction models under different training data sizes. Sensitivity analyses are performed for model hyperparameters and their tuning methods to adapt for other datasets. The results show that LingoTrip outperforms data-driven models in terms of prediction accuracy, transferability (between individuals), zero/few shot learning (limited training sample size) and interpretability of predictions. The LingoTrip model can facilitate the effective provision of personalized information for system crowding and disruption contexts (i.e., proactively providing information to targeted individuals).http://www.sciencedirect.com/science/article/pii/S1077291X25000025Large Language ModelsIndividual MobilitySpatiotemporal Context PromptPersonalied InfromationPublic Transport |
| spellingShingle | Zhenlin Qin Pengfei Zhang Leizhen Wang Zhenliang Ma LingoTrip: Spatiotemporal context prompt driven large language model for individual trip prediction Journal of Public Transportation Large Language Models Individual Mobility Spatiotemporal Context Prompt Personalied Infromation Public Transport |
| title | LingoTrip: Spatiotemporal context prompt driven large language model for individual trip prediction |
| title_full | LingoTrip: Spatiotemporal context prompt driven large language model for individual trip prediction |
| title_fullStr | LingoTrip: Spatiotemporal context prompt driven large language model for individual trip prediction |
| title_full_unstemmed | LingoTrip: Spatiotemporal context prompt driven large language model for individual trip prediction |
| title_short | LingoTrip: Spatiotemporal context prompt driven large language model for individual trip prediction |
| title_sort | lingotrip spatiotemporal context prompt driven large language model for individual trip prediction |
| topic | Large Language Models Individual Mobility Spatiotemporal Context Prompt Personalied Infromation Public Transport |
| url | http://www.sciencedirect.com/science/article/pii/S1077291X25000025 |
| work_keys_str_mv | AT zhenlinqin lingotripspatiotemporalcontextpromptdrivenlargelanguagemodelforindividualtripprediction AT pengfeizhang lingotripspatiotemporalcontextpromptdrivenlargelanguagemodelforindividualtripprediction AT leizhenwang lingotripspatiotemporalcontextpromptdrivenlargelanguagemodelforindividualtripprediction AT zhenliangma lingotripspatiotemporalcontextpromptdrivenlargelanguagemodelforindividualtripprediction |