A comparative study of machine translated subtitles based on the user-centered approach: a case study between Bilibili and YouTube

Abstract Bilibili has become a famous video-sharing platform in China, attracting a growing number of visitors. Expect for the videos made by uploaders themselves, an increasing number of videos in different languages are uploaded. This year, the platform announced that automatic translation can be used for free to offer an alternative solution to the language barrier. Previous research has been done to test the impact machine-translated subtitles of TED (Technology, Entertainment, and Design) or MOOCs (Massive Open Online Courses) have on the users’ reception. Very little attention has been paid to do a comparative study between different platforms. This research did a pilot study to compare machine translated subtitles on Bilibili (MTS-B) with those on YouTube (MTS-Y). Based on the human evaluation of the quality and the questionnaire conducted online, the author analyzed the relationship between the quality, users’ comprehension and attitude towards machine translated subtitles. The results show that the quality has little impact on the users’ comprehension on the video. However, when MTS-Y performs better in accuracy, participants had a better attitude towards it.