Querying Vague Spatial Objects in Databases with Vasa

Recent years have been witness to the increasing efforts of scientists to design concepts and implementations that can adequately handle the vagueness and imprecision that is widespread in spatial data. Plenty of spatial entities, especially those that occur naturally such as mountains and biotopes, contain intrinsic vague attributes that make their representation as crisply bounded entities ineffective and far from exact. Other examples of such spatial objects include population density, pollution clouds, oil pouches, and even lakes and rivers whose water levels are not determinate but rather can change depending on the pluvial activity. In the context of spatial databases, retrieval and handling of vague spatial objects through querying is critical in exploiting the functionality of the database. Thus, it is important that query mechanisms are able to handle the vagueness that is included in the data models used to represent the objects that are stored. In this paper we present a Vague Spatial Algebra (VASA) capable of handling spatial vagueness. VASA is defined on the basis of exact models for crisp spatial objects and includes the definitions of vague spatial data types as well as the operations and predicates needed to effectively manipulate instances of these data types. Using VASA as a basis, we introduce the appropriate concepts for enabling database querying that can help exploit the power of the vague spatial data model.