도시정비사업 승소결정요인에 관한 계량법학적 연구 : 인공신경망과 의사결정나무 분석기법을 활용한*

Abstract

This study analyzes the winning determinants for lawsuit cases in urban and housing redevelopment projects based on jurimetrics utilizing data mining methodologies of decision tree analysis and artificial neural networks. According to the decision tree analysis, for the prerequisite defensive process, it is noteworthy that ‘legal relation in the past’, ‘no cooperative membership or ownership’, and ‘no inherent authorization defect’ are significant in CHAID analysis, and ‘legal relation in the past’ and ‘no cooperative membership or ownership’ are significant in CRT analysis. The CRT analysis for the actual action shows that ‘defect in relation with articles of association’ and ‘no cooperative membership or ownership’ are substantial. The artificial neural network analysis shows that the variables of which importance of normalization is more than 50% are ‘legal relation in the past’, and ‘no cooperative membership status or ownership’ in the prerequisite defensive process. This study intends to help the urban and housing redevelopment projects participants predict the litigation results which can prevent unproductive legal disputes.