| In the age of big data,storage device and computing power which are continuously improving have made people have the inexhaustible data resource and the analysis methods which are results-oriented,People are neither to set the overall any longer,nor to according to the sampling plan design to sample in the overall,instead of all the data.Therefore there is an increasing amount of people are beginning to neglect sampling,they think sampling will make no sense in the age of big data.However,sampling technique have been acknowledged since the end of 19 th century and is continued to this day,it not only plays a role in access to data and statistical inference but also represents an idea which make big data small and remains the same data structure.In order to recognize the reasons for look down upon sampling in the age of big data more clearly,first started from the historical background of the origin of sampling,noted that emergence of sampling is for the purpose of representing overall data with small data,stated its value in surveys,considered the characteristics of the age of big data and shortcomings of sampling itself,suggests that,in age of big data,the threat which the sampling faced is to be replaced.To illustrate sampling still exists necessarily in the age of big data,analysis of the concept of big data,overall,large sample and gross data,indicates that big data is not overall and it cannot represent the gross data from point of the concept and definition,in addition,analyzed the difference of characteristics,resources data analysis methods and ideas between big data and sampling data.At last but not least,to retain sample or even want to make sample play a greater value in the age of big data can be studied from two aspects,on the one hand,take the integration of sampling data and big data into account,on the other hand,regard random and non-random sampling as data-processing techniques applied to big data analytical methods and processes. |
