eKoNLPy: Korean NLP Python Library for Economic Analysis

Overview

eKoNLPy is an open-source Korean NLP library built specifically for analyzing economic and financial texts. While general-purpose Korean NLP tools exist, they often struggle with the specialized vocabulary of central banking, finance, and economic policy. eKoNLPy fills this gap by extending the popular MeCab morphological analyzer with domain-specific dictionaries and sentiment classification tools.

The library was originally developed to support research on Bank of Korea monetary policy communications and has since grown into a versatile toolkit for anyone working with Korean economic texts.

What It Does

eKoNLPy’s core feature is its hawkish-dovish classifier — a sentiment analysis tool that can determine whether a given monetary policy statement leans toward tightening (hawkish) or easing (dovish). This is particularly useful for researchers, analysts, and journalists tracking central bank policy signals.

The library also provides enhanced morphological analysis that correctly handles Korean financial terminology, company names, and institutional terms that standard NLP tools often misparse.

Technical Details

eKoNLPy extends the MeCab-ko tokenizer with curated dictionaries of economic terms, and provides pre-trained sentiment models for monetary policy text classification. It is built as a Python package with a simple API that integrates naturally into data science workflows using pandas and scikit-learn.

The library is available on PyPI and the source code is hosted on GitHub under the MIT license.

Related Research

eKoNLPy served as a foundational tool for several publications on central bank communication analysis, including research with the Bank of Korea on evaluating monetary policy communication effectiveness.

Collaborators

• Bank of Korea: Domain expertise and data for monetary policy analysis
• Open Source Community: Ongoing development and contributions