부동산시장 소비자심리지수의 효과적 통계정보 전달을 위한 가시화에 관한 연구

This study applied a variety of multi-variate analysis method and investigated a possibility of Chernoff Face graph as an effective statistical visualization tool for Real estate Consumer Psychology Index (RCPI). The findings of this study are as following. Firstly, this paper shows the possibility of applying Chernoff Face to RCPI on time series and regional comparison which makes it clearly recognize differences, although Chernoff Face slightly changes by an arbitrary parameter setting. It was also suggested that excessive many comparisons of Chernoff Face in a graph may reduce the ability of recognition and effectiveness. Secondly, an interesting pattern was detected on two dimensional space of MDS (Multi-Dimensional Scaling) where a count-clock wise orientation on t ime series of RCPI was c ircled. This pattern can be interpreted as a result of reflection of real estate market’s characteristic which has an economic cycle: a bull and bear market on horizontal axis. Thirdly, it was proposed that RCPI can be effectively visualized by clusters in terms of linking Chernoff Face to Cluster Analysis. As a result, each Chernoff Face illustrates each cluster’s characteristics usefully where resulting clusters were categorized by Clustering Analysis. Fourthly, it was presented that regional differences can be effectively compared through a combination of GIS mapping and Chernoff Face. At this time, a comparison should be made on the level of metropolitan council, not local government due to recognition ability.