home_page: where the user put in the url for the channel he/she is interested in
piechart: sentimental analysis
wordcloud: word frequency
A week ago, in UTEK (Engineering Kompetition) debate session, one of the debate questions was whether lethal threats on social media platform should be banned. This inspired me if we can create some kind of web app to evaluate the emotional implication in comments. This will help social media to better report unsuitable content; it will also help individuals such as youtubers to analyze massive number of comments.
What it does
- take in a URL of a Youtube channel
- analyze sentimental contents in comments into negative, positive, and neutral categories
- present data visually with pie chart, flow chart and word cloud (unimplemented)
Challenges I ran into
- time limit
- too cold to sleep
What I learned
- time / project management matters!
- how to use java (I have never learnt about java before)
What's next for AnalyzeYoutubeComment
- combine backend and front end
- implement the website
- combine with the Youtube back end
- more personal customs (analyze videos over a period of time)
- add example comments(top three replied comments, top three thumb-up comments)
- export as .pdf
- share to Google Drive, Evernote etc.