Salvato in:
| Autori principali: | , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | https://arxiv.org/abs/2509.19135 |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866914052401266688 |
|---|---|
| author | Luo, Wenying Lin, Zhiyuan Xu, Wenhao Liu, Minghao Li, Zhi |
| author_facet | Luo, Wenying Lin, Zhiyuan Xu, Wenhao Liu, Minghao Li, Zhi |
| contents | Human mobility traces, often recorded as sequences of check-ins, provide a unique window into both short-term visiting patterns and persistent lifestyle regularities. In this work we introduce GSTM-HMU, a generative spatio-temporal framework designed to advance mobility analysis by explicitly modeling the semantic and temporal complexity of human movement. The framework consists of four key innovations. First, a Spatio-Temporal Concept Encoder (STCE) integrates geographic location, POI category semantics, and periodic temporal rhythms into unified vector representations. Second, a Cognitive Trajectory Memory (CTM) adaptively filters historical visits, emphasizing recent and behaviorally salient events in order to capture user intent more effectively. Third, a Lifestyle Concept Bank (LCB) contributes structured human preference cues, such as activity types and lifestyle patterns, to enhance interpretability and personalization. Finally, task-oriented generative heads transform the learned representations into predictions for multiple downstream tasks. We conduct extensive experiments on four widely used real-world datasets, including Gowalla, WeePlace, Brightkite, and FourSquare, and evaluate performance on three benchmark tasks: next-location prediction, trajectory-user identification, and time estimation. The results demonstrate consistent and substantial improvements over strong baselines, confirming the effectiveness of GSTM-HMU in extracting semantic regularities from complex mobility data. Beyond raw performance gains, our findings also suggest that generative modeling provides a promising foundation for building more robust, interpretable, and generalizable systems for human mobility intelligence. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_19135 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | GSTM-HMU: Generative Spatio-Temporal Modeling for Human Mobility Understanding Luo, Wenying Lin, Zhiyuan Xu, Wenhao Liu, Minghao Li, Zhi Machine Learning Artificial Intelligence Human mobility traces, often recorded as sequences of check-ins, provide a unique window into both short-term visiting patterns and persistent lifestyle regularities. In this work we introduce GSTM-HMU, a generative spatio-temporal framework designed to advance mobility analysis by explicitly modeling the semantic and temporal complexity of human movement. The framework consists of four key innovations. First, a Spatio-Temporal Concept Encoder (STCE) integrates geographic location, POI category semantics, and periodic temporal rhythms into unified vector representations. Second, a Cognitive Trajectory Memory (CTM) adaptively filters historical visits, emphasizing recent and behaviorally salient events in order to capture user intent more effectively. Third, a Lifestyle Concept Bank (LCB) contributes structured human preference cues, such as activity types and lifestyle patterns, to enhance interpretability and personalization. Finally, task-oriented generative heads transform the learned representations into predictions for multiple downstream tasks. We conduct extensive experiments on four widely used real-world datasets, including Gowalla, WeePlace, Brightkite, and FourSquare, and evaluate performance on three benchmark tasks: next-location prediction, trajectory-user identification, and time estimation. The results demonstrate consistent and substantial improvements over strong baselines, confirming the effectiveness of GSTM-HMU in extracting semantic regularities from complex mobility data. Beyond raw performance gains, our findings also suggest that generative modeling provides a promising foundation for building more robust, interpretable, and generalizable systems for human mobility intelligence. |
| title | GSTM-HMU: Generative Spatio-Temporal Modeling for Human Mobility Understanding |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2509.19135 |