Large-scale Geospatial Analytics: Problems, Challenges, and Opportunities

Tsz Nam Chan, Leong Hou U, Byron Choi, Jianliang Xu, Reynold Cheng

Abstract: Geospatial analytics is an important field in many communities, including crime science, transportation science, epidemiology, ecology, and urban planning. However, with the rapid growth of big geospatial data, most of the commonly used geospatial analytic tools are not efficient (or even feasible) to support large-scale datasets. As such, domain experts have raised the concerns about the inefficiency issues for using these tools. In this tutorial, we aim to arouse the attention of database researchers for this important, emerging, database-related, and interdisciplinary topic, which consists of four parts. In the first part, we will discuss different problems and highlight the challenges for two types of geospatial analytic tools, which are (1) hotspot detection and (2) correlation analysis. In the second and third parts, we will specifically discuss two geospatial analytic tools, namely kernel density visualization (the representative hotspot detection method) and K-function (the representative correlation analysis method), respectively, and their variants. In the fourth part, we will highlight the future opportunities for this topic.