3 ms·
Hunting Down Twitter Sh’tposters with Node.js/Puppeteer
Hey HN!
My name is Jared Wright, yesterday I did a live stream where I built a web scraper to find the funniest Twitter accounts (sh!tposters) based on a topic. Seeing as the recommendation feed is eh. It took sub 2 hours to finish the script. Seeing how Twitter is closing their API a bit I also wanted to do some web scraping to show alternatives to getting data.
Here’s the live stream for anyone interested https://www.youtube.com/live/-Hfx9tCeShA?feature=share
I started streaming 2 weeks ago to educate people on some topics around how to build things on your own (finetuning LLMs, web scraping, chrome extensions, etc) rather than being dependent on APIs all the time. Also just like having fun hacking at night and need someone to talk too…Anyway! Have fun and ask any question below I’m here all day.
- jawerty 3y agoLink to the live stream for anyone interested https://www.youtube.com/live/-Hfx9tCeShA?feature=share https://www.youtube.com/live/-Hfx9tCeShA?feature=share Also feel free to message me for any questions on how to web scrape and learn browser automation in general and how to use these processes to build your own datasets
- Rational444444 3y ago[flagged]