Smart City Gnosys
Smart city article details
| Title | Castle: A Context-Aware Spatial-Temporal Location Embedding Pre-Training Model For Next Location Prediction |
|---|---|
| ID_Doc | 13471 |
| Authors | Cheng J.; Huang J.; Zhang X. |
| Year | 2023 |
| Published | International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences - ISPRS Archives, 48, 4/W2-2022 |
| DOI | http://dx.doi.org/10.5194/isprs-archives-XLVIII-4-W2-2022-15-2023 |
| Abstract | Next location prediction is helpful for service recommendation, public safety, intelligent transportation, and other location-based applications. Existing location prediction methods usually use sparse check-in trajectories and require massive historical data to capture complex spatial-temporal correlations. High spatial-temporal resolution trajectories have rich information. However, obtaining personal trajectories with long time series and high spatiotemporal resolution usually proves challenging. Herein, this paper proposes a two-stage Context-Aware Spatial-Temporal Location Embedding (CASTLE) model, a multi-modal pre-training model for sequence-to-sequence prediction tasks. The method is built in two steps. First, large-scale location datasets, which are sparse but easier to be acquired (i.e., check-in and anomalous navigation data), are used for pre-training location embedding to capture the multi-functional properties under different contexts. After that, the learned contextual embedding is used for downstream location prediction in small-scale but higher spatiotemporal resolution trajectory datasets. Specifically, the CASTLE model combines Bidirectional and Auto-Regressive Transformers to generate contextual embedding vectors rather than a fixed vector for each location. Furthermore, we introduce a location and time-aware encoder to reflect the spatial distances between locations and visit times. Experiments are conducted on two real trajectory datasets. The results show that the CASTLE model can pre-train beneficial location embedding and outperforms the model without pre-training by 4.6-7.1%. The proposed method is expected to improve the next location prediction accuracy without massive historical data, which will greatly drive the use of trajectory data. © Author(s) 2023. |
| Author Keywords | Geospatial Data; Location Embedding; Location Prediction; Smart City; Trajectory Mining; Ubiquitous Computing |
Similar Articles
| Id | Similarity | Authors | Title | Published |
|---|---|---|---|---|
38438  | 0.892 | Zhang R.; Guo J.; Jiang H.; Xie P.; Wang C. | Multi-Task Learning For Location Prediction With Deep Multi-Model Ensembles | Proceedings - 21st IEEE International Conference on High Performance Computing and Communications, 17th IEEE International Conference on Smart City and 5th IEEE International Conference on Data Science and Systems, HPCC/SmartCity/DSS 2019 (2019) |
| 39208 | 0.882 | Li P.; Wang Z.; Zhang X.; Wang P.; Liu K. | Next Arrival And Destination Prediction Via Spatiotemporal Embedding With Urban Geography And Human Mobility Data | Mathematics, 13, 5 (2025) |
| 15923 | 0.863 | Tsiligkaridis A.; Zhang J.; Paschalidis I.C.; Taguchi H.; Sakajo S.; Nikovski D. | Context-Aware Destination And Time-To-Destination Prediction Using Machine Learning | ISC2 2022 - 8th IEEE International Smart Cities Conference (2022) |
| 11877 | 0.859 | Li X.; Hu R.; Wang Z. | Beyond Fixed Time And Space: Next Poi Recommendation Via Multi-Grained Context And Correlation | Neural Computing and Applications, 35, 1 (2023) |
| 25830 | 0.851 | Liu C.; Zhang H.; Zhu G.; Guan H.; Kwong S. | Exploring Trajectory Embedding Via Spatial-Temporal Propagation For Dynamic Region Representations | Information Sciences, 668 (2024) |