This study analyzes the political slants of user comments on Korean partisan media. We built a BERT-based classifier to detect political leaning of short comments via the use of semi-unsupervised deep learning methods that produced an F1 score of 0.83. As a result of classifying 27.1K comments, we found the high presence of conservative bias on both conservative and liberal news outlets. Moreover, this study discloses a considerable overlap of commenters across the partisan spectrum such that the majority of liberals (88.8%) and conservatives (63.7%) comment not only on news stories resonating with their political perspectives but also on those challenging their viewpoints. These findings advance the current understanding of online echo chambers.