Reproduce “MobTCast: Leveraging Auxiliary Trajectory Forecasting for Human Mobility Prediction” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Hao Xue, Flora D.Salim, Yongli Ren, Nuria Oliver
MobTCast presents a deep learning model for next Place-of-Interest (POI) prediction that comprehensively integrates temporal, semantic, social, and geographical contexts. It utilizes an auxiliary trajectory forecasting task and a novel consistency loss to enhance prediction accuracy, outperforming prior state-of-the-art methods by an average of 7.22% across multiple datasets.
Key insights
- Integrating temporal, semantic, social, and geographical contexts comprehensively within a unified deep learning framework significantly enhances next POI prediction accuracy.
- An auxiliary trajectory forecasting task, coupled with a novel consistency loss, effectively models geographical context and improves the main POI classification task by enforcing coherence between predicted location and predicted POI.
- The model's ability to infer social links from check-in patterns demonstrates robustness and generalizability to datasets without explicit social network information.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.