<abstract xmlns="http://www.w3.org/1999/xhtml"><p>Selection is a key element of the cartographic generalisation process, often being its first stage. On the other hand it is a component of other generalisation operators, such as simplification. One of the approaches used in generalization is the condition-action approach. The author uses a condition-action approach based on three types of rough logics (Rough Set Theory (RST), Dominance-Based Rough Set Theory (DRST) and Fuzzy-Rough Set Theory (FRST)), checking the possibility of their use in the process of selecting topographic objects (buildings, roads, rivers) and comparing the obtained results. The complexity of the decision system (the number of rules and their conditions) and its effectiveness are assessed, both in terms of quantity and quality – through visual assessment. The conducted research indicates the advantage of the DRST and RST approaches (with the CN2 algorithm) due to the quality of the obtained selection, the greater simplicity of the decision system, and better refined IT tools enabling the use of these systems. At this stage, the FRST approach, which is characterised by the highest complexity of created rules and the worst selection results, is not recommended. Particular approaches have limitations resulting from the need to select appropriate measurement scales for the attributes used in them. Special attention should be paid to the selection of network objects, in which the use of only a condition-action approach, without maintaining consistency of the network, may not produce the desired results. Unlike approaches based on classical logic, rough approaches allow the use of incomplete or contradictory information. The proposed tools can (in their current form) find an auxiliary use in the selection of topographic objects, and potentially also in other generalisation operators.</p></abstract>

Selection is a key element of the cartographic generalisation process, often being its first stage. On the other hand it is a component of other generalisation operators, such as simplification. One of the approaches used in generalization is the condition-action approach. The author uses a condition-action approach based on three types of rough logics (Rough Set Theory (RST), Dominance-Based Rough Set Theory (DRST) and Fuzzy-Rough Set Theory (FRST)), checking the possibility of their use in the process of selecting topographic objects (buildings, roads, rivers) and comparing the obtained results. The complexity of the decision system (the number of rules and their conditions) and its effectiveness are assessed, both in terms of quantity and quality – through visual assessment. The conducted research indicates the advantage of the DRST and RST approaches (with the CN2 algorithm) due to the quality of the obtained selection, the greater simplicity of the decision system, and better refined IT tools enabling the use of these systems. At this stage, the FRST approach, which is characterised by the highest complexity of created rules and the worst selection results, is not recommended. Particular approaches have limitations resulting from the need to select appropriate measurement scales for the attributes used in them. Special attention should be paid to the selection of network objects, in which the use of only a condition-action approach, without maintaining consistency of the network, may not produce the desired results. Unlike approaches based on classical logic, rough approaches allow the use of incomplete or contradictory information. The proposed tools can (in their current form) find an auxiliary use in the selection of topographic objects, and potentially also in other generalisation operators.

The use of rough rules in the selection of topographic objects for generalizing geographical information

Polish Cartographical Review

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.

Selection is a key element of the cartographic generalisation process, often being its first stage. On the other hand it is a component of other generalisation...

The use of rough rules in the selection of topographic objects for generalizing geographical information

Publicado en línea: 06 jul 2020

Páginas: 1 - 15

Recibido: 02 mar 2020

Aceptado: 25 mar 2020

DOI: https://doi.org/10.2478/pcr-2020-0001

Palabras clave
cartographic generalization, Rough Set Theory, ominance-Based Rough Set Theory, Fuzzy-Rough Set Theory, decision rules

© 2020 Anna Fiedukowicz, published by Sciendo

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.

The use of rough rules in the selection of topographic objects for generalizing geographical information

Publicado en línea: 06 jul 2020

Páginas: 1 - 15

Recibido: 02 mar 2020

Aceptado: 25 mar 2020

DOI: https://doi.org/10.2478/pcr-2020-0001

Palabras clavecartographic generalization, Rough Set Theory, ominance-Based Rough Set Theory, Fuzzy-Rough Set Theory, decision rules

© 2020 Anna Fiedukowicz, published by Sciendo

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.

Palabras clave
cartographic generalization, Rough Set Theory, ominance-Based Rough Set Theory, Fuzzy-Rough Set Theory, decision rules