
The Bank of Russia intends to analyze economic activity in the Russian Federation on the basis of news. The Central Bank has developed a method of processing large data arrays (Big Data) using indicators of the dynamics of economic activity, and not just the data of official statistics, the Prime Agency reports.
The news indicator built by specialists of the Central Bank showed good leading properties, allowing you to quickly evaluate the current situation in the economy before the publication of official statistics. It is built on the basis of daily analysis of news sites using machine learning and allows you to quickly assess the situation in the economy before the publication of official statistics. The indicator is "copyright", about its official status in the report of the Central Bank "Assessment of economic activity based on text analysis" is not about.
The results of the work will serve as an important starting point for the further development of analytical indicators using Big Data, the Bank of Russia claims.
As initial data, news articles taken from Internet resources are used. Analysis of news articles is carried out by the methods of text analysis and machine learning, which are developed relatively recently, but quite quickly gained widespread in scientific work, including economic, indicated in the Central Bank.
“This is due to the fact that the news is not only the main source of information, with their help you can also recognize the mood of journalists and respondents interviewed regarding the current situation and convert them into quantitative data,” the regulator’s website explains.
Today, many economists are already developing analytical indicators based on the Big Data analysis, this direction is already actively developing in a number of foreign central banks. The demand for such indicators is due to the need to obtain the most operational and complete information for making decisions in the context of a rapidly changing situation in the modern economy, added to the Central Bank.