LDA 토픽모형을 이용한 부동산시장 뉴스 토픽에 관한 연구

Research Purpose: The purpose of this study is to analyze the trends in the real estate market by identifying the main key words and topics in online newspaper. Research Methods: The data used for analysis is that collected monthly from November 11, 2021 to November 5, 2022 in relation to real estate market before and after the 20th Government. The analysis method used word frequency-inverse document frequency(TF-IDF) and LDA topic model. Results in Research: Before the new government the dominant topic was housing supply related to reconstruction and development. Meanwhile, macroeconomic topics such as interest rates and inflation, which are considered major factors affecting housing prices, appeared relatively less dominant. However, after the new government macroeconomic and finance topics became dominant. Interest rates, jeonse(a unique Korean rental system), and price declines, which showed low frequency before the new government, showed high frequency after the new government. On the other hand, housing supply related to reconstruction and development, appeared relatively less dominant. Research Conclusion: Before and After the new government, trends in the real estate market appeared differently despite the short period. Considering that the real estate market is inefficient compared to other assets, it can be inferred that it is responding relatively sensitively despite the short time series period.