A Stochastic Viewpoint on the Generation of Spatiotemporal Datasets

Abstract

The issue of standardized generation scheme of spatio- temporal datasets is a research area of growing importance. In case of the lack of large real datasets, especially, benchmarking spatio-temporal database requires the generation of synthetic datasets simulating the real-word behavior of spatial objects that move and evolve over time. Recently, a few studies have been conducted on the generation of artificial datasets from a different point of view. For more realistic datasets, this paper proposes a novel framework, called state-based movement framework (SMF) to provide more generalized framework for both describing and generating the movement of complexly moving objects which simulate the movement of real-life objects. Based on Markov chain model, a well-known stochastic model, the proposed model classifies the whole trajectory of a moving object into a set of movement state. From some illustrative examples, we show that the proposed scheme is able to generate various realistic datasets with respect to the given input parameters.

This work was done as a part of Information and Communication Fundamental Technology Research Program, supported by Ministry of Information and Communication in Republic of Korea.