search for


A news-based topic modeling of wildfire over suppression time
Journal of the Korean Data & Information Science Society 2023;34:245-54
Published online March 31, 2023;
© 2023 Korean Data and Information Science Society.

Haebin Lim1 · Dongyeob Kim2 · Sanghoo Yoon3

13Department of Data Science, Daegu University
2Department of Forest Resources, Daegu University
Correspondence to: 1 Undergraduate student, Department of Data Science, Daegu University, Gyeongbuk 38453, Korea.
2 Assistant professor, Department of Horticulture, Daegu University, Gyeongbuk 38453, Korea.
3 Associate professor, Department of Data Science, Daegu University, Gyeongbuk 38453, Korea. E-mail:
Received January 9, 2023; Revised January 27, 2023; Accepted January 30, 2023.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License ( which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
As Korea is a forest country with many mountainous areas, a large wildfire is caused by a dry atmospheric environment and wind. large-scale wildfires causing enormous economic losses. It affects livestock, forest products, and homes in addition to reducing biodiversity and destroying wildlife habitats. This study conducted network analysis and topic modeling to see how wildfire-related information spreads through the news when a large-scale wildfire occurs. 9,224 articles were collected and analyzed with the keywords ‘Goseong wildfire’ and ‘Uljin wildfire’ through Bigkinds. 13 topics were selected by the minimum perplexity, and ordered as weather, crime, donation, community support, cause, celebrity donation, broadcast, extinguishment, politics, incident, special disaster area, disaster support, and damage situation. The cause of the wildfire was the main concern of the news in the case of the relatively short wildfire in Goseong, but public interest in donations and goods to help the victims was high in the case of the Uljin wildfire. In the case of the Goseong wildfire, wildfire information was spreading centering on government and political news, but in the case of the Uljin wildfire, news information was spreading centering on the relief association.
Keywords : Disaster, network analysis, topic modeling, wildfire.