This paper focused on a method of data cleansing for used condominium database that is open to the public by the Ministry of Land, Infrastructure, Transport and Tourism. By identifying and removing error records, which are not transaction outliers, it enables us to analyze this dataset appropriately. In this study, we proposed two methods to deal with serious error, that is suspect of “Fat finger error”. In addition, as a result of noise removal, the analysis results using the hedonic model have improved significantly.
