온라인 커뮤니티에 나타난 부동산시장 참여자 심리와 부동산시장 변동 간의 인과관계분석

Consumer Sentiment in On-line Community and the Variation of Real Estate Market

This paper conducted an empirical analysis how the psychological changes of consumers effects the price and transaction volume in the real estate market. Methodologically, we utilized the social media(on-line community) data, which is popularly used for the survey of citizens satisfaction and preference in marketing, consumer services, politics and elsewhere. We crawled postings uploaded to the real estate social media community for seven months from August 28, 2017 to April 9, 2018, using text mining programming language R. The collected data were quantified through sentiment analysis. To compare and analyze the sentiment with the actual real estate market, we used the existing real estate market indexes, such as housing transaction volume, rental transaction index, trading price index, and rental price index. Assuming that any change in consumer’s sentiment that occurs prior to an event (i.e, changes in the real actual estate market) can be used as an evidence of causation, we applied Granger Causality test. The results revealed that the patterns in consumer sentiment appear repeatedly in trading index after approximately nine weeks, and that consumer sentiment has a significant effect on real estate market. Finally, scrutinizing on-line postings and predicting the future with sentiment analysis could help create a sustainable and stable real estate market.