Bank of Russia develops big data-based leading indicator

The Bank of Russia presents the first research dedicated to big data analysis. The ‘Text Mining-based Economic Activity Estimates’ research paper introduces the method for calculating the leading indicator of Russian economic activity based on a daily online news mining involving machine learning.

The global information environment has been changing rapidly in recent years. This comes along with increasing volumes of information and the development of new data processing methods which open space for analysis of various aspects of people’s lives, including economic issues.

Many economists are already developing analytical indicators based on big data. They are used in many different fields, which include both economic forecasting and real-time analysis. Such indicators are in demand because of the need to receive as much real-time and complete information as possible to make a decision in a rapidly changing modern economy. This makes some foreign central banks engage in big data-based analysis.

The news indicator elaborated by the author of the article released today demonstrated good forward-looking features that allow real-time assessment of the current state of the economy before the official statistics are released. The outcomes will be a starting point for further development of big data-based analytical indicators in the Bank of Russia.

The views expressed in the paper are solely those of the author and do not necessarily represent the official position of the Bank of Russia.