The interpretation of topographic maps is a complex task which requires training and expertise in the cartographic domain. The paper presents CLASSIAT, an intelligent system for classifying environmental morphologies from digitised maps. The knowledge is automatically acquired and updated usign a symbolic sysitem for incrementally learning from positive and negative examples. CLASSIAT, which has been designed for urban planning purposes, is organised in three processing phases: pre-processing, learning, and classification. Results of the application of a prliminary release of the system for the recognition of land patterns from digitised maps have shown a classification accuracy of about 98%, confirming the appropriateness of a symbolic learning approach to map interpretation.
The Application of Machine Learning Techniques to Map Interpretation
ESPOSITO, Floriana;LANZA, Antonietta
1996-01-01
Abstract
The interpretation of topographic maps is a complex task which requires training and expertise in the cartographic domain. The paper presents CLASSIAT, an intelligent system for classifying environmental morphologies from digitised maps. The knowledge is automatically acquired and updated usign a symbolic sysitem for incrementally learning from positive and negative examples. CLASSIAT, which has been designed for urban planning purposes, is organised in three processing phases: pre-processing, learning, and classification. Results of the application of a prliminary release of the system for the recognition of land patterns from digitised maps have shown a classification accuracy of about 98%, confirming the appropriateness of a symbolic learning approach to map interpretation.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.