Page 375 - Special Topic Session (STS) - Volume 2

STS500 Li J.
and appropriate conclusions can be drawn through the reciprocal
recombination of data between different fields; in terms of the way of thinking,
the perception of useless data should be changed as any data in the big data
contains the “information” and the information that seems “wrong” is just the
evidence for “correct” information. The value of big data requires frequent
mining and use from different angles, so the potential value will be released
continuously.

(III) Analyzing the application space of big data in the business process
of government statistics
Currently, the application of big data in government statistics mainly
focuses on specific fields, and the government statistical methods are
perfected by technical means. For example, in the population statistics, the big
data from mobile communications is used to realize the dynamic monitoring
of population; in the agricultural statistics, the remote sensing technology is
applied to obtain the information about crop planting area, floor area of
facility agriculture and yield prediction of crops, etc.; in the real estate price
statistics, the data endorsed on the internet is used in most of the large and
medium-sized cities to calculate the price index of new houses and second-
hand houses; in the transportation statistics, the big data generated from the
highway network monitoring system is used to reckon the highway freight
information.
A complete set of business processes has been formed in practice for
government statistics over the years, which is mainly divided into three links,
namely data collection, data analysis and data issuing. The whole business
system is quite mature and basically meets the social statistical needs.
Therefore, it is thought in this paper that the application of big data in
government statistics should not be limited to the level of technical means
and also the specific fields of statistics under the current business process
system of statistics and a deeper application is that the thinking of big data
should run through the government statistics.
By taking the statistical business process as an entry point, the experience
and methods are drawn from the application of big data in enterprises to
realize the perfection of data collection, data analysis and data issuing based
on the analysis on expandable application space of big data in each link,
improve the data quality for government statistics, deepen the data analysis
and enhance the foresight of trend analysis.
In the data collection process, a large amount of source data is manually
collected, the data quality is interfered by human factors, and most of data is
collected after the event. Therefore, such data has poor timeliness to reflect
the hot spots and emergencies and low predictability for future trend. It is
hereby proposed to optimize the data collection methods by means of big

364 | I S I W S C 2 0 1 9