Uma análise experimental de modelos de relacionamentos direcionais em banco de dados espaciais

Abstract

The use of spatial data is constantly growing in several areas of knowledge. This information is stored in Spatial Databases, which use approximations to simplify objects and thus reduce execution time. Spatial queries employ spatial relations, such as directional relationships. There are a number of models proposed in the literature to identify directional relationships, such as the simple and detailed Direction-Relation Matrix (DRM) and the Objects Interaction Matrix (OIM), that do not have practical implementations. Therefore, application developers who want to use directional relationships need to implement their own versions of these models, which is expensive, time-consuming task and can lead to errors. Thus, this work proposes implementations of the DRM and OIM models in the R language, also performing experimental analysis to evaluate and compare these models. According to the analysis performed, using a dataset with approximately 100,000 spatial objects from the state of São Paulo, the simple DRM model proved to be the fastest since it uses less complex spatial operations.