Application Bases and General Ideas of Big Data in Price Statistics
As scanner data and online transaction data are becoming more common,the popularity of big data in price statistics has also increased.Based on the application of scanner data,this paper sorts out basic works of using big data in the price statistics,and then discusses related issues of basic price,and then systematically considers the price index theory,pointing out that the long-term index generated through the short-term bilateral index chain on the basis of dynamic data set may have chain drift due to the high frequency and high chum of big data,so that the long-term price changes aren't available for accu-rate measurement,thus clarifies the applicability and rationality of the multilateral index or the use of a dynamic method to generate a index in the big data environment.Finally,considering the characteristics of big data set,the advantages of dividing homogeneous items,the application presumption of unit value method,and combined with statistical practice and the overall grasp of index theories and methods,we can get inspirations of big data application for the basic works side,basic price side,in-dex compilation side,the comprehensive index(e.g.CPI)compilation side respectively.
big dataunit value methodmultilateral indexdynamic methodchain drift
NETL
NSTL
万方数据
