CrossSim: Toward Cross-System Trajectory Similarity Computation via Representation Learning

Zijian Cao, Dong Zhao, Xiyuan Dong, Qiyue Wang, Haitao Yuan, Yujia Gao, Huadong Ma

Abstract: Trajectory similarity computation is essential for various downstream applications, such as anomaly route detection, order matching, and digital contact tracing. However, its effectiveness is confined within a single system due to privacy concerns associated with sharing raw trajectories across different systems. In this article, we propose CrossSim, a novel framework designed to efficiently retrieve similar trajectories across all systems while preserving individual privacy. Our framework comprises three main components: 1) a Trajectory Encoding Model that transforms trajectories into high-quality representations, where similarity relationships are reflected by their distances; 2) a two-stage optimization mechanism, including a Contrastive Similarity Learning stage and a Federated Similarity Learning stage, that alleviates the impact of heterogeneous similarity relationships across different systems on model training without aggregating raw trajectories; and 3) a Similar Trajectory Retrieval procedure that obtains top-k similar trajectories from all systems without sharing raw trajectories. We conduct comprehensive experiments on three real-world datasets to evaluate the effectiveness of our proposed framework. The evaluation results demonstrate that CrossSim outperforms all existing schemes.

External IDs:dblp:journals/iotj/CaoZDWYGM25