Designing a Cartographic Ontology for Use with Expert Systems

The democratization of mapping represents a positive trend for society, however within this movement, little focus to cartographic quality has been paid. The result is often poorly presented maps which fail to convey the information intended. Software packages offer very little, if any, guidance to the user when it comes to offering cartographic assistance such as symbolization recommendations. Additionally, users often do not read software manuals, and, instead, opt to solve the problem by trial and error or ask an available expert. The combination of scarce cartographic documentation and the resistance of users to use documentation present an opportunity for expert systems to provide better assistance through a non-passive help system. The paper presents the development of ontologies and their potential contribution to the construction and successful execution of an expert system. Traditionally, expert systems learned the situation through a dialog with the user after which it would infer an expert answer. A problem with this model is that if the users do not have sufficient knowledge of cartography to answer these questions, the expert system will be less effective at helping. If the expert system can understand the situation without querying the user, it can make proactive inferences, thus, reducing the burden on the user and reducing the amount of time spent interacting with the expert system. To effectively replace user input to an expert system, it must be able to understand the concept of a map. The construction of an ontology will provide this understanding.