Assessments and forecasts of housing markets can provide insight into the fundamental sustainability of housing and construction. The home sales index (HSI) is considered one of the most important factors for forecasting economic trends of housing markets in the real estate and construction industry, and researchers have tried to develop relevant forecasting models for the HSI. The autoregressive integrated moving average (ARIMA) has generally been used for forecasting future trends based on time series but without investigating any of the influences of social factors. However, there are many demands for effective HSI forecasting by identifying the various social factors influencing the HSI. The HSI can be effectively forecasted in advance by observing several social factors. Such forecasting methods can be developed using big data analytic methods that focus on the relationship between those factors and HSI using web search data. This study suggests a methodology for forecasting model development with the provision of fundamental attributes and the pros and cons of each model to which the multiple regression analysis (MRA) and the artificial neural network (ANN) were applied. The forecasting performance of these models was compared with that by ARIMA. This study also quantifies the HSI forecasting accuracy between MRA and ANN based on social factors obtained from web search data. The forecast HSI values using ARIMA are more accurate than those of MRA and ANN. The lowest mean absolute error and normalized mean-square error for each model were calculated as 1.680 and 1.089 by MRA, 1.557 and 1.843 by ANN, and 0.173 and 0.294 by ARIMA, respectively. This methodology could allow many researchers to create and develop forecasting models using web search data for HSI forecasting and other related economic indexes.
|Journal||Journal of Management in Engineering|
|Publication status||Published - 2018 Mar 1|
Bibliographical noteFunding Information:
This work was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) and
funded by the Ministry of Education of the Korean government (NRF-2015R1D1A1A01058221). The authors gratefully acknowledge this support.
© 2017 American Society of Civil Engineers.
All Science Journal Classification (ASJC) codes
- Building and Construction
- Industrial relations
- Strategy and Management
- Management Science and Operations Research