Abstract:
In this paper, we study the disparity of sentiments sentiment of actors in a video
and its respective user comments. We propose a method to calculate the sentiment
score for each video and each comment user. This method enables us to place the
video sentiment and comment sentiment into the same time period to explore the
sentiment factor. We adopt the correlation coe cient between the sentiment scores
of actors and viewers to measure the in
uence. We categorize the youtube videos
with respect to subscriber count and try comment on the fact how it a ects the
correlation between the video and comment sentiment. Community detection and
machine learning are integrated into our approach. We nd that the di erence for
correlation coe cients exists between di erent levels of the number of subscribers
and their audience base.
