An analysis of online educational videos in social media based on verbal content

Nowadays educational videos constitute one of the most popular online forms of educational material in social networks. The study of the popularity of the videos and the language characteristics of the speakers are the two main research axes of the present paper. It starts with the selection of the educational videos from the YouTube and it suggests he evaluation formula of their popularity considering both the positive and the negative opinions of the users, who have watched them. The videos are classified based on this formula into two categories: the most popular and the least popular. Then the language characteristics of the speakers are extracted from the transcript of every video and a clustering procedure follows, which presents the differences between the language characteristics of the most popular videos and the least popular videos. The findings have shown that the sentence length used by the speakers and the pace of speech affect the user´s choice to watch a video or not.