The research on news clustering for Korean news articles based on specific issues or topics is very few. This study proposed a method of constructing a correct answer data set for monthly issues for 1 years worth of Korean news (approximately 4.58 million articles) and the appropriate preprocessing method and clustering algorithm for effective clustering of this news were researched. The clustering results were visualized and the differences among the different methods were comparatively analyzed.
Jeong-Soo Kim, Yuchul Jung, Heung-Seon Oh, Kwangyoung Kim and Jungsun Yoon. A Study on Clustering News Articles in Korean.
DOI: https://doi.org/10.36478/jeasci.2019.3492.3497
URL: https://www.makhillpublications.co/view-article/1816-949x/jeasci.2019.3492.3497