About Ambiguities in Visual GIS Query Languages: a Taxonomy and Solutions

Abstract

In visual query languages, especially in Geographic Information Systems (GIS), a query can lead to multiple interpretations for the system and for the user. An origin is that a unique working space is used to express different kinds of information. Another origin is that the user gives his own semantics to information. Ambiguities represent a big challenge in visual query languages because they are found at different levels and sublevels. Ambiguities depend on the action and on the interpretation of the user and on the representation of the system. We propose a taxonomy of ambiguities in visual GIS query languages in order to determine which case of ambiguity is met and at which level. We also propose different solutions to resolve the ambiguities we detected. A model has been made to solve a particular case of ambiguity in the Lvis [3] visual language. This model shows that the number of semantics in a spatial ambiguity can be very high. An extension of Lvis is currently implemented to test some resolution ways.