Evaluation of Comment Sentiments in Educational YouTube Videos

Official URL: https://risc01.sabanciuniv.edu/record=b3205823

In the digital era, YouTube has become one of the most popular resource for online education. A prior study has reported a bias towards male narrators in YouTube search results for educational videos. Videos with male narrators are exposed more on the platform for both in STEM(Science, Technology, Engineering, and Mathematics) and NON-STEM subjects. To further investigate the bias, the user generated comments are used in this thesis. To better understand viewers’ interactions with YouTube videos, it is crucial to explore the sentiment patterns in the comments. Moreover, user engagements differ between genres. This thesis aims to explore the sentiment expressed in comments on educational YouTube videos. The research has two objectives. Firstly, understanding the sentiment behavior of users towards male and female narrators. Secondly, investigating the sentiment pattern between different subjects: STEM and NON-STEM. In the first part of the study, video’s ranking on YouTube platform is taken into consideration to further investigate the behavioral change between the videos with different rankings. In the second part, the comment’s ranking is taken in to account to understand the comment ranking behavior. For both of the parts the comment sentiment patterns are examined and the behavior is compared between perceived genders of the narrators and subjects. By addressing these objectives, this thesis aims to understand the underlying sentiment behavior behind comments and the differences between the perceived genders of video narrators and subjects in the context of educational content on YouTube.