Review and Perspective for Distance Based Trajectory Clustering

TL;DR

This paper reviews existing trajectory clustering distances, introduces a new Symmetrized Segment-Path Distance (SSPD), and compares its effectiveness with other distances using hierarchical and affinity propagation clustering methods.

Contribution

The paper provides a comprehensive review of trajectory distances and proposes the novel SSPD, improving clustering performance in geolocalized data analysis.

Findings

SSPD outperforms existing distances in clustering accuracy

Comparison shows SSPD yields more coherent trajectory clusters

Hierarchical and affinity propagation methods benefit from the new distance

Abstract

In this paper we tackle the issue of clustering trajectories of geolocalized observations. Using clustering technics based on the choice of a distance between the observations, we first provide a comprehensive review of the different distances used in the literature to compare trajectories. Then based on the limitations of these methods, we introduce a new distance : Symmetrized Segment-Path Distance (SSPD). We finally compare this new distance to the others according to their corresponding clustering results obtained using both hierarchical clustering and affinity propagation methods.