How do you go where? Improving next location prediction by learning travel mode information using transformers

TL;DR

This paper introduces a transformer-based neural network that incorporates travel mode information and an auxiliary task to significantly improve the accuracy of next location prediction in human mobility analysis.

Contribution

It proposes a novel transformer decoder model that learns travel mode as an auxiliary task, enhancing next location prediction accuracy over state-of-the-art methods.

Findings

Significant improvement in F1-score over existing methods.

Travel mode information and temporal features positively impact prediction accuracy.

Performance varies notably with different travel modes.

Abstract

Predicting the next visited location of an individual is a key problem in human mobility analysis, as it is required for the personalization and optimization of sustainable transport options. Here, we propose a transformer decoder-based neural network to predict the next location an individual will visit based on historical locations, time, and travel modes, which are behaviour dimensions often overlooked in previous work. In particular, the prediction of the next travel mode is designed as an auxiliary task to help guide the network's learning. For evaluation, we apply this approach to two large-scale and long-term GPS tracking datasets involving more than 600 individuals. Our experiments show that the proposed method significantly outperforms other state-of-the-art next location prediction methods by a large margin (8.05% and 5.60% relative increase in F1-score for the two datasets, respectively). We conduct an extensive ablation study that quantifies the influence of considering temporal features, travel mode information, and the auxiliary task on the prediction results. Moreover, we experimentally determine the performance upper bound when including the next mode prediction in our model. Finally, our analysis indicates that the performance of location prediction varies significantly with the chosen next travel mode by the individual. These results show potential for a more systematic consideration of additional dimensions of travel behaviour in human mobility prediction tasks. The source code of our model and experiments is available at https://github.com/mie-lab/location-mode-prediction.