A Logical Framework for Frequent Pattern Discovery in Spatial Data

Inrecenttimes,severalextensionsofdataminingmethods andtechniqueshavebeenexploredaimingatdealingwith advanceddatabases.Manypromisingapplicationsof inductivelogicprogramming(ILP)toknowledgediscovery indatabaseshavealsoemergedinordertobenefitfrom semanticsandinferencerulesoffirst-orderlogic.Inthis paper,anILPframeworkforfrequentpatterndiscoveryin spatialdataispresented.Thepatterndiscoveryalgorithm operatesonfirst-orderlogicdescriptionscomputedbyan initialstepoffeatureextractionfromaspatialdatabase.The algorithmbenefitsoftheavailablebackgroundknowledge onthespatialdomainandsystematicallyexploresthe hierarchical structure of task-relevantgeographiclayers. Preliminary results have beenobtained by running the algorithmSPADAonspatial data froman Italian province.

• Type: Conference
• Year: 2001
• Where: FLAIRS
• Authors: Donato Malerba, Floriana Esposito, Francesca A. Lisi
• Keywords: Artificial Intelligence, FLAIRS 2001, Froman Italian Province, Hierarchical Structure, Semanticsandinferencerulesoffirst-orderlogic.Inthis Paper