摘要
房地产作为国民经济的支柱产业,其价格预测受到学者的广泛关注。现有的预测研究,其数据存在严重的滞后性,影响预测的有效性。互联网搜索引擎关键词在表征用户信息需求、行为趋势等方面的能力日益显著,为分析用户消费信息需求、消费行为趋势等提供了较高质量的实时数据。本文在对搜索行为和商品房价格决定机制的理论分析基础上,论证了将关键词关注度指数加入回归预测模型的合理性,并使用Google Trend关键词数据进行了实证研究,结果显示这一改进可以提高对商品房价格指数的预测能力,降低预测误差。
Real estate is a pillar industry of the national economy, of which the price forecast is widely concerned by scholars. The previous studies of housing price prediction have serious lag in data sources, thus the validity of the fore-cast is restricted. The popularity of the Internet enables the increasingly significant capacity of search engine keywords in the characterization of users’ real information needs and behavior trends. This provides real-time data for scholars to analyze consumers’ information demand and behavior. In this paper we carry out a systematic theoretical analysis on search behavior and housing price determination mechanism. On this basis, we demonstrate the rationality of adding keywords attention index into the housing price prediction model and conduct an empirical research using Google Trend Keywords data. The results illustrate that the new model can improve the predictive power and accuracy of housing price prediction and reduce the prediction error.
出处
《预测》
CSSCI
北大核心
2015年第4期65-70,共6页
Forecasting
关键词
搜索引擎
关键词
商品房价格指数
房价预测
搜索行为
search engine
keywords
commercial housing price index
housing price forecast
search behavior