A Methodology for Real-Time Spatiotemporal Data-Intensive Computation

We propose a methodology for real-time spatiotemporal data-intensive computation. The proposed methodology allows researchers to implement spatiotemporal data-intensive tasks on any platform, including mobile, laptop, desktop, and high performance computing. The methodology comprises of three major components: efficient analytics, predictive analytics, and meta-analytics. One of the main features of the proposed methodology is that by taking into account the unique characteristics of any spatiotemporal dataset, the available computing platform, and the required response time, it can provide optimal, near-optimal, or at a minimum, an acceptable solution for computing the task with least compromise on accuracy. Experimental results on real data show that the methodology is able to compute within available time constraint and computing platform.