Construction andApplication ofBayesian Network ModelforSpatial DataMining

Theadvent ofspatial information technologies, suchas GIS,GPS andRemoteSensing, havegreatenhancedour capabilities tocollect andcapture spatial data. How todiscover potentially useful information andknowledge frommassive amountsofspatial dataisbecoming acrucial project forspatial analysis andspatial decision making. Bayesian networks havea powerful ability forreasoning andsemantic representation, whichcombinedwithqualitative analysis and quantitative analysis, withprior knowledge andobserved data, andprovides aneffective waytospatial datamining. Thispaperfocuses on construction andlearning aBayesian network modelforspatial datamining. Firstly, we introduce thetheory ofspatial data mininganddiscuss thecharacteristics ofBayesian networks. A framework andprocess ofspatial datamining isproposed. Then weconstruct aBayesian network modelforspatial datamining withthegiven dataset. Theexperimental results demonstrate the feasibility andpractical oftheproposed approach tospatial data mining. Finally, wedrawaconclusion andshowfurther avenues forresearch. Keywords-Bayesian networks; spatial datamining; knowledge