Here is a little video I made of insights that can be found by using the Youtube API. I learned how to use it from the youtube channel StrataScratch. I really enjoyed the way they showed not only how to use the API, but also how to bring the info into a pandas data frame and how to use good coding fundamentals to order your code. So, I decided to make my first YouTube API project for them. Here it is.
This video explains all the things I found out about StrataScratch YouTube subscribers/commenters:
What did I find?
Using the Youtube API I scraped all the comments on the StrataScratch channel. Then I scraped the most recent actions of each commenter. Usually, an action would be uploading a video or subscribing to a channel.
Next, I Grouped the data to see what channels were most subscribed to by the people who had commented as you can see here:
After fiddling with a few other tables for context I came to my next question:
Who are the biggest StrataScratch fans?
Assuming it would be those who had commented the most, I put them in order. Two things to note. 1. Stratascratch is actually the top commenter which makes sense. 2. I've blurred out the lines with a neat little bit of code because it seemed a little intrusive to have the channel names visible:
NLP?
I was just coming off of another NLP project for someone who had scraped a bunch of Upwork posts so I figured hey, why not use the NLP code and spaCy to see if there are some words used most frequently that could be useful?
In the end, there were many more questions, but from here I have created a list of data that might have insights.
-How could I find the niche youtube channels that StrataScratch commenters watch?
-What video data do the StrataScratch competitor channels have in common? What are the outlier videos like? etc....