Learning context-aware region similarity with effective spatial normalization over Point-of-Interest data

Jiahui Jin, Yifan Song, Dong Kan, Binjie Zhang, Yan Lyu, Jinghui Zhang, Hongru Lu

Abstract: Highlights•A CARE model is designed to learn region similarity with spatial and task contexts.•A spatial normalization technique is proposed to quantify region importance to neighbors.•The performance of CARE has been shown through experiments on real-world datasets.