- Volume 19 Issue 2
This study analyzes the factors that influence commercial gentrification in Seoul by using both logit model analysis and machine learning with data cumulated from 2015 to 2018 regarding 158 market areas. Logit analysis indicates that log(market area average monthly rent) and the ratio of the purchasing amount by customers aged 40 and younger to total sales in the restaurant and retail business category are statistically significant at 1%; the increase in sales per female customer aged between 30 and 39 in the restaurant and retail business category is statistically significant at 5%; and the increase in number of retailers with a business history of less than two years in the franchise business category is significant at 10%. Machine learning indicates that significant factors ordered by importance are the total retail area, the existence of an industrial complex within the market area, the existence of a traditional market within the market area, the location of subway stations within the market area, the increase of entertainment facilities such as movie theaters within the market area, average monthly rent, and the growth rate of average monthly rent. The contribution of this research is threefold. First, this study analyzes the entire commercial area of Seoul, Korea. Second, this study provides a foundation for future research on predictive indicators by empirically investigating the factors that influence commercial gentrification in Seoul. Lastly, this study introduces various methods of research by utilizing a machine learning approach.
Commercial Area;Gentrification;Logit Analysis;Machine Learning