LSTM-Based Deep Learning Model for Predicting Individual Mobility Traces of Short-Term Foreign Tourists

Abstract

The increasing availability of trajectory recordings has led to the mining of a massive amount of historical track data, allowing for a better understanding of travel behaviors by revealing meaningful motion patterns. In the context of human mobility analysis, the problem of motion prediction assumes a central role and is beneficial for a wide range of applications, including for touristic purposes, such as personalized services or targeted recommendations, and sustainability studies related to crowd management and resource redistribution. This paper tackles a particular case of the trajectory prediction problem, focusing on large-scale mobility traces of short-term foreign tourists. These sparse trajectories, short and non-repetitive, lack spatial and temporal regularity, making prediction analysis based on individual historical motion data unreliable. To face this issue, we hereby propose a deep learning-based approach, taking into account the collective mobility of tourists over the territory. The underlying semantics of motion patterns are captured by means of a long short-term memory (LSTM) neural network model trained on pre-processed location sequences, aiming to predict the next visited place in the trajectory. We tested the methodology on a real-world big dataset, demonstrating its higher feasibility with respect to traditional approaches.

LSTM-Based Path Prediction for Effective Sensor Filtering in Sensor Registry System

“…Recurrent neural networks (RNN) are effective tools for processing sequential data and can be applied to path prediction as well [ 31 , 32 ]. Crivellari et al propose a series of methods to analyze call detail records related to tourists’ behavior in Italy, such as geo-embedding [ 33 , 34 ], predicting individual mobility traces [ 35 ], trajectory translation [ 36 ], and urban traffic forecasting [ 37 ]. Similar to natural language processing, these methods have three significant components [ 38 ]: 1.…”

Forecasting Spatially-Distributed Urban Traffic Volumes via Multi-Target LSTM-Based Neural Network Regressor

“…LSTM [40] is a complex recurrent module, composed of four distinctive neural networks interacting between each other. Due to the characteristic capability of handling sequential data, it has been utilized in a variety of applications dealing with motion trajectory prediction [41][42][43] and mobility-related time series forecasting [44][45][46]. A few adoptions to studies on traffic analysis are also present, but has been mainly modeled for singularly predicting flows on a selected target location [47][48][49][50].…”

Sustainable Determinants That Affect Tourist Arrival Forecasting

“…Therefore, the statistical data should be collected so that the tourist arrivals are differentiated as individual or group visits, which is not the case now. Or, even better, by their arrival status, e.g., whether they arrived independently or in a group and whether a group was organised by a third party [73,74]. It is of enormous importance to have such data to study the effects and the relationships between variables.…”