DOI QR코드

DOI QR Code

Prediction of Housing Price Index using Data Mining and Learning Techniques

데이터마이닝과 학습기법을 이용한 부동산가격지수 예측

  • Received : 2021.05.28
  • Accepted : 2021.08.20
  • Published : 2021.08.28

Abstract

With increasing interest in the 4th industrial revolution, data-driven scientific methodologies have developed. However, there are limitations of data collection in the real estate field of research. In addition, as the public becomes more knowledgeable about the real estate market, the qualitative sentiment comes to play a bigger role in the real estate market. Therefore, we propose a method to collect quantitative data that reflects sentiment using text mining and k-means algorithms, rather than the existing source data, and to predict the direction of housing index through artificial neural network learning based on the collected data. Data from 2012 to 2019 is set as the training period and 2020 as the prediction period. It is expected that this study will contribute to the utilization of scientific methods such as artificial neural networks rather than the use of the classical methodology for real estate market participants in their decision making process.

4차 산업에 대한 관심이 증폭되면서 데이터를 활용한 과학적 방법론이 발전하고 있지만 부동산 분야에 대한 연구는 데이터 수집의 한계점을 내포하고 있다. 더불어 일반 시장 참여자들의 지식이 확장되면서 정성적인 심리가 부동산 시장에 미치는 영향이 커지고 있다. 때문에 본 연구에서는 기존의 원천 데이터가 아닌 심리적 부분을 반영한 정량 데이터를 텍스트마이닝과 k-meas 알고리즘을 통해 수집하는 방안을 제안하고 수집된 데이터를 바탕으로 인공신경망 학습을 통해 주택 지수의 방향성을 예측하고자 한다. 2012년부터 2019년까지의 데이터를 학습 기간으로 하고 2020년도를 예측 기간으로 설정하여 실험을 진행한 결과, 두 가지 CASE에서 예측 능력이 약 80% 이상으로 우수하였고 주택지수의 상승 구간에서의 예측 강도 또한 우수한 결과를 보였다. 본 연구를 통해서 의사결정에 있어서 부동산 시장 참여자들에게 인공신경망과 같은 과학적 방식의 활용도 증가 및 고전적 방식에서 벗어난 원천 데이터의 대체 데이터 확보 등에 대한 노력이 증진되기를 기대한다.

Keywords

References

  1. J. Y. Lee & J. P. Ryu. (2021). Prediction of Housing Price Index Using Artificial Neural Network. Journal of the Korea Academia-Industrial, 22(4), 228-234. DOI : 10.5762/KAIS.2021.22.4.228
  2. D. W. Kim & J. S. Yu. (2013). Analysis on How Psychological Attitudes on the House Price Affect the Trading Volume. Korean Association For Housing Policy Studies, 21(2), 73-92. DOI : G704-000825.2013.21.2.005
  3. K. M. Kim. (2018). A Study on Dynamic Correlations between the Seoul Apartment Market and Factors of Macroeconomic Variables. Korea Real Estate Academy, 73(1), 115-129.
  4. E. M. Kim, S. B. Kim & E. S. Cho. (2020). Using Mechanical Learning Analysis of Determinants of Housing Sales and Establishment of Forecasting Model. Journal of Cadastre & Land Informatix, 50(1), 181-201. DOI : 10.22640/lxsiri.2020.50.1.181
  5. H. J. Chun. (2020). Big Data Time Series Analysis of the Relationship Between News Coverage Apartment Price and Trading Volume. Korean Society of Real Estate Law Society, 24(2), 53-69. DOI : 10.32989/rel.2020.24.2.53
  6. M. J. Noh & S. J. Yoo. (2016). A Relationship between Sales Prices of APT and Consumer Sentiment. Korea Research Institute for Human Settlements, 89(1), 3-13. DOI : 10.15793/kspr.2016.89..001
  7. H. S. Yoo & J. H. Chung. (2015). The Lead-Lag Relationship between Housing PurchasePrice Index and Consumer Sentiment Index. korea real estate research institute, 25(4), 49-61. UCI : G704-001886.2015.25.4.002
  8. T. J. Cho. (2014). A Study on the effect of the sentiment index to the housing market. Korean Association For Housing Policy Studies, 22(3), 25-48. UCI : G704-000825.2014.22.3.002
  9. D. W. Kim & J. S. Yu. (2013). An Analysis on How Psychological Attitudes on the House Price Affect the Trading Volume. Korean Association For Housing Policy Studies, 21(2), 73-92. UCI : G704-000825.2013.21.2.005
  10. S. Y. Heo, J. Y. Kim & T. H. Moon. (2019). Consumer Sentiment in On-line Community and the Variation of Real Estate Market. Journal of the Korean housing association, 17(4), 31-41. DOI : 10.22313/reik.2019.17.4.31
  11. J. P. Ryu, C. H. Han & H. J. Shin. (2016). Sector Investment strategies Using Big Data Trends. Korea Institute of Enterprise Architecture, 13(1), 111-121. DOI : G704-SER000010357.2016.13.1.004
  12. J. P. Ryu, H. J. Shin, M. H. Kim & J. K. Baek. (2017). Pattern Analysis of Stock Prices Using Machine Learning and Data Visualization. Korea Institute of Enterprise Architecture, 14(2), 189-197.
  13. G. Neha & R. Rinkle. (2017). Analysis and visualization of Twitter data using k-means clustering. ICICCS, 15(1), 15-16. DOI : G704-SER000010357.2016.13.1.004
  14. J. P. Ryu & H. J. Shin. (2018). Portfolio Selection Strategy Using Deep Learning. Korea Institute of Enterprise Architecture, 15(1), 43-50. DOI : 10.22865/jita.2018.15.1.43
  15. J. M. Lim & M. H. Lim. (2016). A Study of the Relationship Between Agents Sentiment and Housing Market. Journal of the Korean regional development association, 28(3), 147-164. UCI : G704-000688.2016.28.3.005 https://doi.org/10.22885/KRDA.2016.28.3.147