A Prediction Study of Real Estate Price Index Based on Web Search Data

摘要：

The web search data which reflects millions of internet users concerns and interests,as well as the trends of their behavior,provides essential data basis for the study of macro-economic issues. Based on the theory of demand and supply,this paper establishes a concept frame from the perspective of the factors which affect housing sales prices,it also reveals that there is a certain correlation and lead-lag relationship between web search data and price index of real estate sales. Empirical results further confirm this relationship is statistically significant The model is able to obtain a better fit when web search data is considered. Model fitting is about 0.979 and the mean absolute percentage error is 1.46％. Simultaneously,the model has very strong time effectiveness comparing with traditional monitor methods,its forecast result can be obtained about one month ahead of the State Statistical Bureaus report

关键词： web search data price prediction price index of real estate sales housing sales prices

作者: Wang Bian Peng Geng Yuan Qingyu Lv Benfu

作者单位: Graduate School of Chinese Academy of Sciences,Beijing,100080

会议类型: 国际会议

会议名称: The Tenth Wuhan International Conference on E-Business(第十届武汉电子商务国际会议)

会议地点: 成都

会议语种:英文

页码: 303-311

在线出版日期: 2011-05-13（万方平台首次上网日期，不代表论文的发表时间）