포털 사이트 내 언론사의 구독자 정치성향 분석을 통한 포털 사이트의 정치적 중립성 제고 방안

Due to the development of the Internet and popularization of smart phones, the proportion of portal sites in Korea has already surpassed traditional media. Accordingly, the composition of political neutrality is also increasing. This is because portal sites expose news on a specific basis. This can make it difficult for subscribers to freely post opinions and accept various opinions on specific issues. Therefore, this study selected two representative portal sites in Korea (Naver, Daum), and used deep learning to grasp the political tendencies of the comments posted in the articles to raise the issue of the portal's current news editing and exposure standards and to overcome them. As a result of research, there is a difference in the political tendencies of each portal site, and the higher the number of “like” comments, the more likely it is the same as the political tendencies of the portal site.