Trajectory Test-Train Overlap in Next-Location Prediction Datasets

TL;DR

This paper investigates the extent of trajectory overlap in next-location prediction datasets, revealing a memorization issue and proposing a reranking method that significantly enhances generalization, especially for unseen trajectories.

Contribution

It uncovers the prevalence of trajectory overlap in datasets and introduces a reranking approach to improve predictor generalization in next-location prediction.

Findings

High trajectory overlap limits predictor generalization.

Proposed reranking improves accuracy up to 96.15% on low-overlap data.

Predictors tend to memorize trajectories rather than generalize.

Abstract

Next-location prediction, consisting of forecasting a user's location given their historical trajectories, has important implications in several fields, such as urban planning, geo-marketing, and disease spreading. Several predictors have been proposed in the last few years to address it, including last-generation ones based on deep learning. This paper tests the generalization capability of these predictors on public mobility datasets, stratifying the datasets by whether the trajectories in the test set also appear fully or partially in the training set. We consistently discover a severe problem of trajectory overlapping in all analyzed datasets, highlighting that predictors memorize trajectories while having limited generalization capacities. We thus propose a methodology to rerank the outputs of the next-location predictors based on spatial mobility patterns. With these techniques, we significantly improve the predictors' generalization capability, with a relative improvement on the accuracy up to 96.15% on the trajectories that cannot be memorized (i.e., low overlap with the training set).