Synthetic and Real Spatiotemporal Datasets

In the context of a spatiotemporal research environment, it is very important to be able to systematically generate data with predictable characteristics. For instance, it allows one to use the same datasets, or others similarly characterized, for benchmarking access structures or mining techniques. This paper presents a survey of existing generators of synthetic spatiotemporal data. It also covers a few real datasets, which are (at the time of this writing) publicly available for research use.