This api takes a youtube video url and returns a table of de-identified comments and their sentiment. This shows the overall sentiment if a youtube video's sentiment is positive, negative or neutral.
This application was built for the Pytorch 2021 hackathon and can be tested at this site.
To run this project, you will need to add the following environment variables to your .env file
DEVELOPER_KEY
Get the youtube api key from your google developer console
GET /api/v1/analyzeComments| Parameter | Type | Description |
|---|---|---|
video |
string |
Required. Url of the video |
Takes a video url and returns the sentiment predictions of each comment on the video.
I continued to firm up my htmx with flask to create a "SPA" like application.
- Support for adding multiple video urls and queing to get video comment sentiments
- Adding support for multiple sites that have comments