Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Self-supervised Trajectory Representation Learning with Temporal Regularities and Travel Semantics

Jiawei Jiang, Dayan Pan, Houxing Ren, Xiaohan Jiang, Chao Li, Jingyuan Wang

TL;DR

A novel Self-supervised trajectory representation learning framework with TemporAl Regularities and Travel semantics, namely START is proposed, which is verified by extensive experiments on two large-scale real-world datasets for three downstream tasks.

Abstract

Trajectory Representation Learning (TRL) is a powerful tool for spatial-temporal data analysis and management. TRL aims to convert complicated raw trajectories into low-dimensional representation vectors, which can be applied to various downstream tasks, such as trajectory classification, clustering, and similarity computation. Existing TRL works usually treat trajectories as ordinary sequence data, while some important spatial-temporal characteristics, such as temporal regularities and travel semantics, are not fully exploited. To fill this gap, we propose a novel Self-supervised trajectory representation learning framework with TemporAl Regularities and Travel semantics, namely START. The proposed method consists of two stages. The first stage is a Trajectory Pattern-Enhanced Graph Attention Network (TPE-GAT), which converts the road network features and travel semantics into representation vectors of road segments. The second stage is a Time-Aware Trajectory Encoder (TAT-Enc), which encodes representation vectors of road segments in the same trajectory as a trajectory representation vector, meanwhile incorporating temporal regularities with the trajectory representation. Moreover, we also design two self-supervised tasks, i.e., span-masked trajectory recovery and trajectory contrastive learning, to introduce spatial-temporal characteristics of trajectories into the training process of our START framework. The effectiveness of the proposed method is verified by extensive experiments on two large-scale real-world datasets for three downstream tasks. The experiments also demonstrate that our method can be transferred across different cities to adapt heterogeneous trajectory datasets.

Self-supervised Trajectory Representation Learning with Temporal Regularities and Travel Semantics

TL;DR

A novel Self-supervised trajectory representation learning framework with TemporAl Regularities and Travel semantics, namely START is proposed, which is verified by extensive experiments on two large-scale real-world datasets for three downstream tasks.

Abstract

Trajectory Representation Learning (TRL) is a powerful tool for spatial-temporal data analysis and management. TRL aims to convert complicated raw trajectories into low-dimensional representation vectors, which can be applied to various downstream tasks, such as trajectory classification, clustering, and similarity computation. Existing TRL works usually treat trajectories as ordinary sequence data, while some important spatial-temporal characteristics, such as temporal regularities and travel semantics, are not fully exploited. To fill this gap, we propose a novel Self-supervised trajectory representation learning framework with TemporAl Regularities and Travel semantics, namely START. The proposed method consists of two stages. The first stage is a Trajectory Pattern-Enhanced Graph Attention Network (TPE-GAT), which converts the road network features and travel semantics into representation vectors of road segments. The second stage is a Time-Aware Trajectory Encoder (TAT-Enc), which encodes representation vectors of road segments in the same trajectory as a trajectory representation vector, meanwhile incorporating temporal regularities with the trajectory representation. Moreover, we also design two self-supervised tasks, i.e., span-masked trajectory recovery and trajectory contrastive learning, to introduce spatial-temporal characteristics of trajectories into the training process of our START framework. The effectiveness of the proposed method is verified by extensive experiments on two large-scale real-world datasets for three downstream tasks. The experiments also demonstrate that our method can be transferred across different cities to adapt heterogeneous trajectory datasets.
Paper Structure (43 sections, 17 equations, 10 figures, 3 tables)

This paper contains 43 sections, 17 equations, 10 figures, 3 tables.

Table of Contents

  1. Introduction
  2. PRELIMINARIES
  3. Notations and Definitions
  4. Problem Statement
  5. Method
  6. Trajectory Pattern-Enhanced Graph Attention Layer
  7. Time-Aware Trajectory Encoder Layer
  8. Trajectory Time Pattern Extraction Module
  9. Time Interval-aware Self-attention Module
  10. Trajectory Representation Pooling
  11. Self-supervised Pre-training Tasks
  12. Span-Masked Trajectory Recovery
  13. Trajectory Contrastive Learning
  14. Model Fine-tuning and Downstream Tasks
  15. Trajectory Travel Time Estimation
  16. ...and 28 more sections

Figures (10)

  • Figure 1: Temporal Regularities and Travel Semantics in Trajectories. (Map data $\copyright$ OpenStreetMap contributors, CC BY-SA.)
  • Figure 2: Overall Framework of START.
  • Figure 3: MAPE on BJ Under Different Scenarios.
  • Figure 4: Performance of $k$-nearest Trajectory Search Task When Selection Proportions $p_d$ Vary.
  • Figure 5: Comparision of Top-3 Similar Trajectories Retrieved by START and Trembr. (Map data $\copyright$ OpenStreetMap contributors, CC BY-SA.)
  • ...and 5 more figures

Theorems & Definitions (3)

  • Definition 1: Road Network
  • Definition 2: GPS-based Trajectory
  • Definition 3: Road-network Constrained Trajectory