Analyzing User Comments on YouTube Coding Tutorial Videos

Video coding tutorials enable expert and noviceprogrammers to visually observe real developers write, debug, and execute code. Previous research in this domain has focusedon helping programmers find relevant content in coding tutorialvideos as well as understanding the motivation and needs ofcontent creators. In this paper, we focus on the link connectingprogrammers creating coding videos with their audience. Morespecifically, we analyze user comments on YouTube codingtutorial videos. Our main objective is to help content creators toeffectively understand the needs and concerns of their viewers, thus respond faster to these concerns and deliver higher-qualitycontent. A dataset of 6000 comments sampled from 12 YouTubecoding videos is used to conduct our analysis. Important userquestions and concerns are then automatically classified andsummarized. The results show that Support Vector Machinescan detect useful viewers' comments on coding videos with anaverage accuracy of 77%. The results also show that SumBasic, an extractive frequency-based summarization technique withredundancy control, can sufficiently capture the main concernspresent in viewers' comments.