Self supervised path learning improves travel time and ranking tasks
AutoPRL: Automated Path Representation Learning via a Learnable View Generator
Machine LearningArtificial Intelligence
Summary
Paths traveled on road networks by vehicles can be represented as sequences of road segments, but teaching computers to understand these paths well is hard. The authors propose AutoPRL, a method that creates better, more meaningful views of paths automatically instead of using hand-made tweaks. This system trains more efficiently by carefully comparing good and bad path examples and balancing different learning goals. The team tested AutoPRL on real-world data and found it worked better and faster than previous methods for tasks like estimating travel times and ranking paths.
What this means in practice
- •For transportation planners: Improve travel time forecasts by using automatically learned path features from road networks for better route analysis.
- •For navigation app developers: Rank alternative routes more effectively by integrating efficient self-supervised path representations into routing algorithms.
Authors
Sean Bin Yang, Ying Sun, Yunyao CHeng, Hao Miao, Jilin Hu, Bin Yang, Torben Bach Pedersen, Christian S. Jensen
Abstract
Self-supervised learning has emerged as a powerful paradigm for Path Representation Learning (PRL) in road networks. PRL aims to learn informative representations of paths traversed by moving objects, such as vehicles, where each path consists of a sequence of connected road segments. These representations are essential for various downstream tasks, including travel-time estimation and path ranking. However, existing methods often suffer from ineffective view generation and inefficient training procedures due to their reliance on heuristic augmentations and computationally intensive contrastive learning. To address these issues, we propose AutoPRL, an effective and efficient self-supervised PRL framework that integrates three key components. The Automatic Data Augmentation (AutoDA) module employs variational autoencoders to generate semantically meaningful and diverse feature-level path views by modeling each topology-aware node embedding with a Gaussian latent distribution, thereby mitigating information loss and enhancing representation quality. Second, the Cross-View Decoupled Contrastive (CVDC) learning mechanism improves training efficiency by disentangling the optimization of positive and negative pairs across node- and path-level views, enabling more stable contrastive learning. Third, an Adaptive Co-Training (ACT) module dynamically adjusts uncertainty-based loss weights to balance multiple objectives, ensuring efficient and scalable training. Extensive experiments on three real-world datasets demonstrate that AutoPRL achieves strong and consistent performance in both effectiveness and efficiency across multiple downstream tasks.