Privacy Implication of Location-Based Service: Multi-Class Stochastic User Equilibrium and Incentive Mechanism

Abstract

As a tool to assist traffic guidance and improve service quality, location-based service (LBS) platforms such as route navigation apps rely heavily on the collection and analysis of users’ location/trajectory information, which may evoke privacy concerns. Because of such privacy concerns, users may choose not to provide their information. In certain cases, this may lead to the problem of insufficient data for LBS applications (e.g., travel time estimation). To address this issue, the paper develops a modeling framework to quantify the levels of privacy for mixed user groups and proposes an incentive mechanism to encourage users to provide their location/trajectory information. It is assumed that LBS users have smaller travel time perception error but experience some extra privacy costs compared with the non-LBS users. A bi-level optimal incentive model with stochastic user equilibrium and elastic demand is developed to capture the mixed behavior of multi-class network users. The problem is solved using a meta-heuristic approach combined of genetic algorithm, successive average algorithm, and multiple behavior equilibrium assignment algorithm. The results reveal that the modeling framework can capture the mixed behavior of groups with different privacy levels. The proposed incentive mechanism is able to ensure sufficient data, and simultaneously minimize the required incentive.