3 ms·
I done sentiment analysis of tweets. It literally was as simple as looking for positive words like 'good', 'amazing', 'awesome', 'brilliant', etc and maybe look
by legrande 3y ago
I done sentiment analysis of tweets. It literally was as simple as looking for positive words like 'good', 'amazing', 'awesome', 'brilliant', etc and maybe looking for a smiley-face emoji alongside the tweet, and I found positive sentiment for brands and interacted with the tweets. I was working for a social media company at the time that done brand engagement. Now Twitter is dead we're looking at Bluesky, Nostr, Lemmy, Mastodon, Threads, etc and do our analysis there. It's very astro-turfey, but it works.
- PaulHoule 3y agoI just gotta make “nazi” and “fascist” go away and it would improve my Mastodon feed considerably. Korzybski would be more radical than that and just ban the verb “to be.”. (Then you can describe the racist behavior of X but not say “X is a racist”) I also want something that disappears images that are mainly text, whether it is a T-shirt with a slogan designed to piss people off or a screenshot from some other social media site it is usually crap.