Research

Working Papers

"경제 분석을 위한 텍스트 마이닝 (text mining for economic analysis)" [In Korean]

joint with Soohyon Kim (Bank of Korea), Jhinyoung Shin (Yonsei University) and Ki Young Park (Yonsei University)

We provide detailed description of how text data analysis is done and review series of studies done through text mining. Natural language can be characterized with ambiguity and obscurity compared to structured data. It is hard to retrieve useful information from text data as it carries natural language itself. Text mining or natural language processing is a multi-discipline area of modern technic in which wecan distill and obtain just what we need from text. With the development of AI and machine learning, text mining is becoming one of the high-end technology in various fields of research even in economics. We expect there will be more demand for text data analysis as it will be complementary to traditional structured data and also as a new source of information.

"Measuring Monetary Policy Surprises Using Text Mining: The Case of Korea"

joint with Soohyon Kim (Bank of Korea) and Ki Young Park (Yonsei University)

We propose a novel approach to measure monetary policy shocks using sentiment analysis. We quantify the tones of 24,079 news articles around 152 dates of Monetary Policy Board (MPB) meetings of the Bank of Korea (BOK) from March 2005 to November 2017. We then measure monetary policy surprises using the changes of those tones following monetary policy announcements and estimate the impact of monetary policy surprises on asset prices. Our measure of monetary policy surprises better explains changes in long-term rates, while changes in the Bank of Korea's base rate are more closely associated with changes in short-term rates (maturity of one year less). Our results strongly suggest that using a text mining approach to measure monetary policy surprises sheds light on information related to forward guidance and market expectations on future monetary policy.

Publications

"토픽모델링과 감성분석에 기반한 금통위 의사록 분석 (Interpreting the Minutes of the Monetary Policy Board through Topic modeling and Sentiment Analysis)" [In Korean]

joint with Bohyun Yoon (Kangwon National University)

Journal of the Korean Data Analysis Society, April 2019, Vol. 21, No. 2, pp. 889-900

This study investigates the existence of additional information on the monetary policy board (MPB) minutes of the Bank of Korea (BOK) through two types of text mining techniques that have been rapidly developing, topic modeling and sentiment analysis. This study extracts a total of five topics from the whole minutes using the topic modeling analysis and names them as the economy, the monetary policy, the financial market, the prices and the debt in consideration of the distribution of the words of each topic. The tones of these topics are included in an augmented Taylor rule as additional explanatory variables to analyze the interest rate decision of the central bank. The ordered probit model reveals that tones of the monetary policy, the prices, and the debt have additional information on the current interest rate decision and that the tone of the economy has predictive power for the future interest rate decision. In particular, the fact that there is predictive power in the future has a great significance in that it can form a right expectation for the market participants, and an investment strategy using this information will be possible.

"Deciphering Monetary Policy Board Minutes with Text Mining: The Case of South Korea"

joint with Soohyon Kim (Bank of Korea) and Ki Young Park (Yonsei University)

Korean Economic Review, 2019, Vol. 35, No. 2

We quantify the Monetary Policy Board minutes of the Bank of Korea (BOK) by using text mining. We propose a novel approach that uses a field-specific Korean dictionary and contiguous sequences of words (n-grams) to capture the subtlety of central bank communications. Our text-based indicator helps explain the current and future BOK monetary policy decisions when considering an augmented Taylor rule, suggesting that it contains additional information beyond the currently available macroeconomic variables.In explaining the current and future monetary policy decisions, our indicator remarkably out performs English-based textual classifications, a media-based measure of economic policy uncertainty, and a data-based measure of macroeconomic uncertainty. Our empirical results also emphasize the importance of using a field-specific dictionary and the original Korean text.

Research Projects

Translating Data into Insights: New Technology Enhances ESG Data Analysis and Investing (Mar 2019 - )

Stanford Precourt Institute for Energy
Researchers: In, S.Y. Y. J. Lee and A. Monk

Evaluating the Effectiveness of the BOK Communication using Text Mining (Mar 2018 - Sep 2019)

Economic Research Institute, The Bank of Korea
Researchers: Park, Ki Young and Lee, Young Joon and Kim, Soohyon

Information Extraction and Summarization System using Big Data and Artificial Intelligence (Jun 2017 - Nov 2017)

Sogang University
Researchers: Jung, Yu Sin and Lee, Young Joon and WISEfn

Google Sites

Report abuse