User identification from mobility traces

Geolocation is a powerful source of information through which user patterns can be extracted. User regions-of-interest, along with these patterns, can be used to recognize and imitate user behavior. In this work we develop a methodology for preprocessing location data in order to discover the most r...

Full description

Saved in:
Bibliographic Details
Published in:Journal of ambient intelligence and humanized computing 2023, Vol.14 (1), p.31-40
Main Authors: Salomón, Sergio, Tîrnăucă, Cristina, Duque, Rafael, Montaña, José Luis
Format: Article
Language:English
Subjects:
Accuracy
Artificial Intelligence
Behavior
Cloning
Computational Intelligence
Data mining
Engineering
Global positioning systems
GPS
Mobility
Neural networks
Original Research
Probability distribution
Robotics and Automation
Semantics
Stochastic models
User Interfaces and Human Computer Interaction
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Geolocation is a powerful source of information through which user patterns can be extracted. User regions-of-interest, along with these patterns, can be used to recognize and imitate user behavior. In this work we develop a methodology for preprocessing location data in order to discover the most relevant places the user visits, and we propose a Probabilistic Finite Automaton structure as mobility model. We analyse both location prediction and user identification tasks. Our model is assessed with two evaluation metrics regarding its predictive accuracy and user identification accuracy, and compared against other models.
ISSN:1868-5137
1868-5145
DOI:10.1007/s12652-018-1117-4