Flash

Abstract

The current explosion in spatial data raises the need for efficient spatial analysis tools to extract useful information from such data. Spatial probabilistic graphical modeling (SPGM) is an important class of spatial data analysis that provides efficient probabilistic graphical models for spatial data. Unfortunately, existing SPGM tools are neither generic nor scalable when dealing with big spatial data. In this work, we present Flash ; a framework for generic and scalable spatial probabilistic graphical modeling (SPGM). Flash exploits Markov Logic Networks (MLN) to express SPGM as a set of declarative logical rules. In addition, it provides spatial variations of the scalable RDBMS-based learning and inference techniques of MLN to efficiently perform SPGM predictions. We have evaluated Flash , based on three real spatial analysis applications, and achieved at least two orders of magnitude speed up in learning the modeling parameters over state-of-the-art computational methods.