Research On The Improvement Of Housing Price Prediction By Public Concern

The research on housing price prediction has always been the focus of real estate economics.In recent years,more and more phenomenas prove that the real estate market is not consistance with the traditional economic hypothesis.The interpretation of house price fluctuation could not improved by the macro and micro economic factors of the traditional economic theory.With the prevalence of behavioral economics,it can help the traditional economic theory to explain the fluctuation of house prices from the perspective of the psychological and behavioral aspects of the market subject.The study of behavioral economics on real estate price fluctuations is mostly based on the analysis of behavioral phenomena of microscopic individuals,and lacks objective indicators that can reflect people's behavior and psychology.With the development of Internet big data,people's attention and expectations for the real estate market can be reflected through related search behavior.The search of microscopic subjects on the Internet has caused changes in the degree of attention of related keywords.The purchase behavior in the real estate market has caused changes in housing prices and trading volumes.People will conduct a series of actions such as collecting relevant information,confirming demand,and selecting judgments before the purchase behavior occurs,which forms the time difference between the change of the network keyword and the change of real estate price.Therefore,aiming at the problem of forecasting the price of house prices under traditional economic theory,from the perspective of the combination of traditional economic theories and behavioral economics,the use of network search data to quantify the degree of public concern and explore the relationship between public attention and house price volatility to reflect the public Concerned web keyword effect on improving housing forecasting accuracy.In this paper,we first summarize the related researches on house prices and traditional economics and behavioral economics,and summarize the research on the correlation between online search data and economic activities,and use public concern to reflect cognitive bias,and use network search data to quantify public concern.It also introduces the behavioral economics and its supplement to traditional economic theory.Secondly,by analyzing the macroeconomic factors that affect house prices,we use stepwise regression to build the basic model of Shenzhen second-hand housing price prediction.Then use time difference correlation analysis,lasso regression analysis and principal component analysis to select and synthesize network keywords.Finally,we combine the synthetic data of the public with the basic forecasting model,analyze the model prediction effect after joining the network synthesis data,and analyze the relationship between public attention and house prices.The conclusions of this paper are as follows:(1)Improve the goodness of fit of the predictive model by adding public attention factors,reducing the forecast error.(2)By examining the causality between house prices and online search data,it is found that there is a Granger causality between the two.(3)The price of second-hand housing in Shenzhen is greatly affected by the price and interest rate of new commercial housing,and the change in prices and interest rates of new homes has a relatively rapid impact on the prices of second-hand housing.(4)Changes in second-hand home prices in Shenzhen are more relevant to the national-level provident fund keywords.