English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

In this paper, we study the problem of mobile user profiling, which is acritical component for quantifying users' characteristics in the human mobilitymodeling pipeline. Human mobility is a sequential decision-making processdependent on the users' dynamic interests. With accurate user profiles, thepredictive model can perfectly reproduce users' mobility trajectories. In thereverse direction, once the predictive model can imitate users' mobilitypatterns, the learned user profiles are also optimal. Such intuition motivatesus to propose an imitation-based mobile user profiling framework by exploitingreinforcement learning, in which the agent is trained to precisely imitateusers' mobility patterns for optimal user profiles. Specifically, the proposedframework includes two modules: (1) representation module, which produces statecombining user profiles and spatio-temporal context in real-time; (2) imitationmodule, where Deep Q-network (DQN) imitates the user behavior (action) based onthe state that is produced by the representation module. However, there are twochallenges in running the framework effectively. First, epsilon-greedy strategyin DQN makes use of the exploration-exploitation trade-off by randomly pickactions with the epsilon probability. Such randomness feeds back to therepresentation module, causing the learned user profiles unstable. To solve theproblem, we propose an adversarial training strategy to guarantee therobustness of the representation module. Second, the representation moduleupdates users' profiles in an incremental manner, requiring integrating thetemporal effects of user profiles. Inspired by Long-short Term Memory (LSTM),we introduce a gated mechanism to incorporate new and old user characteristicsinto the user profile.

Reinforced Imitative Graph Representation Learning for Mobile User Profiling: An Adversarial Training Perspective