These written and oral presentations, and all the material contained within, are protected by copyright law.
GEOLIS : Un Système d’information logique pour l’organisation et la recherche de données géolocalisées
2009 January, 22
202 pages
Language: French
Team(s): LIS
Keywords: Information system, geographic data, logic, concept analysis, data exploration, spatial relations, navigation, querying
Summary:
The amount of geographical data available and processed around the world keeps increasing. Thus, accessing this kind of data is a key issue. In this thesis, we propose a new paradigm for geographical data organization and retrieval. Our approach is based on Logical Information System (LIS) and their underlying theory: Logical Concept Analysis. In LIS, the expressiveness of logics enables to describe objects and the logical deduction is used to organize and retrieve data. We propose a new geographical data model centered on the geographical object and that allows to consider flexible sets of geographical data. We define spatial logics to characterize geographical data by their geometry and their spatial relations (topological relations and distance). We also introduce a new way to explore geographical data based on a dynamic combination of querying, navigation and visualization. Our proposal relies on three complementary views over the explored dataset: the query, the selection and the navigation index. The query is a logical formula describing the seek data. The selection is the set of objects described by the query which are visualized in a cartographic area. The navigation index is a synthetic summary of the properties describing each object of the selection. We describe the implementation of the data model and the different views in a prototype that enabled to conduct experiments on two real datasets. Experiments show the benefits of using the tool to get acquainted with a new geographical dataset and to build complex queries using navigation.