Information Access through Conceptual Structures and GIS

This paper presents a new technique for information access based on a combination of Geographic Information Systems (GIS) and conceptual structures as modeled in relational concept analysis. It is based on the idea that traditional relational databases restrict their data types to linearly ordered scales. Since spatial data are not linearly ordered, they cannot be modeled directly in relational databases, but require GIS. A Conceptual Query Interface (CQI) separates the query syntax from conditions for the data types and presents the data in a conceptual environment. It therefore allows a common modeling of linearly ordered and spatial data. Previously CQI has been applied to building a graphical interface for faceted thesauri. The main concern of our modeling of GIS and conceptual structures is not geography, but applications that utilize abstract spaces, such as information spaces, knowledge spaces or the World Wide Web. We present a graphical interface that allows access of spatial, hierarchical, and linear data in a common manner. We discuss its features and limits.