Formatting is a bit off on mobile, but I'm sure you're aware and working on it.
If you haven't thought of it already, you could turn this into a somewhat modular tool. Someone can go to a page, enter the subreddit of their choice, the keywords they want (or just let your software parse the most popular words that aren't common language), and have a nice chart like this quickly.
Would love to hear more on the methodology that drives the chart. Is it raw number of mentions of a stock, or does it use sentiment analysis for positive/negative mentions of a stock?
Didn't a lot of sentiment analysis/NLP companies already tried to make Wallstreetbets sentiments since forever, and couldn't actually make useful data because of the noise (intentional). The only way today is different is that because these days they are united under singular stock, so its pretty easy. But regular wallstreetbets activity, no chance.
That's a really slick website. However it still can't tell the difference between 3 letter words (or 4) that also happen to be symbols. Take for example APPS on the Ticket Sentiment (https://swaggystocks.com/dashboard/wallstreetbets/ticker-sen...). I haven't seen anyone talk about Digital Turbine Inc, but lots of people talking about investing apps. Would be nice if Reddit had some autofill feature that prepended $ to symbols. $APPS vs apps. Perhaps a mod feature they could enable?
No, its my own deduction. I've been in that site for more than a year and there are a lot of noise and spam stock on purpose to obfuscate sentiment from NLP stuffs. It has a lot of satire in it.
Nice chart, can you share some insight, how are building this, I believe you use reddit api and store the comments and run NLP on that? cleaning data is a headache.
Also, "interest" would be more accurate than "sentiment" in its current form.