(image: Korea Bizwire)
SEOUL, Feb. 17 (Korea Bizwire) — Linguistic analysis through artificial intelligence (AI) technology may increase the accuracy of economic research, a central bank report said Thursday.
The Bank of Korea (BOK) compiled big data from 128,000 business assessment reports published by securities analysts and implemented a natural language processing method to analyze the qualitative information included in the reports.
The text analysis came up with an estimate of business performance sorted by the type of industry, which turned out to be useful for making a forecast of macroeconomic indices like gross domestic product (GDP) and business sentiment index (BSI).
Analyzing the business performance index of all industries and changes in leading economic indicators showed a correlation with leading economic indicators not shown on the benchmark Korea Composite Stock Price Index (KOSPI) consensus estimate.
“This indicates that the text information provided by analysts reflects new information that numbers do not provide,” the BOK said.
The report said that text is the most fundamental form of exchanging information, without limits to the range of information it can send. Text analytics technology, therefore, is highly valuable in the field of economics.
The most up-to-date natural language processing platforms such as ChatGPT show that text analytics may bring a new tide of innovation to the field of automated economic analysis.
Kevin Lee (kevinlee@koreabizwire.com)
