Multi View Spatial-Temporal Model for Travel Time Estimation

TL;DR

This paper introduces MVSTM, a novel multi-view spatial-temporal model that effectively captures complex spatial, temporal, and semantic relationships for accurate taxi arrival time prediction in intelligent transportation systems.

Contribution

The paper presents a new multi-view model combining graph2vec, dual-channel temporal modules, and structural embeddings to improve travel time estimation accuracy.

Findings

Outperforms existing methods on large-scale taxi data

Effectively models complex spatial-temporal relationships

Demonstrates superior prediction accuracy

Abstract

Taxi arrival time prediction is essential for building intelligent transportation systems. Traditional prediction methods mainly rely on extracting features from traffic maps, which cannot model complex situations and nonlinear spatial and temporal relationships. Therefore, we propose Multi-View Spatial-Temporal Model (MVSTM) to capture the mutual dependence of spatial-temporal relations and trajectory features. Specifically, we use graph2vec to model the spatial view, dual-channel temporal module to model the trajectory view, and structural embedding to model traffic semantics. Experiments on large-scale taxi trajectory data have shown that our approach is more effective than the existing novel methods. The source code can be found at https://github.com/775269512/SIGSPATIAL-2021-GISCUP-4th-Solution.