Ask HN: Does anyone use Bayesian stats to filter interesting stories on the Net?
There are a lot of stuff online and if you want to read something interesting you need to scroll endlessly on different platforms. Hacker News is one of them (even that it has top stories, you may be not interested in any of them but like something that was submitted but not get that much attention), other are:
* DEV.to
* Twitter Hashtags, Communities and timeline
* Facebook Groups and timeline
So my idea was why not use Bayesian SPAM filter to sort interesting stories. Feed them all in BogoFilter (That I use on my Linux box with Claws-Mail email client) or something similar. I think that most SPAM filter use statistics after Paul Graham famous essay.
You don't even need AI for this. I'm almost certain that your and mine interesting stories have a lot of common that you can train SPAM filter to filter out noise.
From my experience and you probably agree only 1-5% of stuff that you see on above platforms are interesting to me and you.
So the question is does anyone tried something like this? I think that I've read a story about someone use something like this. But I would love to see something that will give me this. It can be Open Source I would welcome that with my open arms. But a paid service like this would be worth my money, even that I don't like subscription based services.
I wanted to share my idea maybe it will inspire someone to create something like this, or maybe something like this already exists.
I have rough idea how something like may work. * You take Open Source RSS Reader and only need to add interface to like or dislike a story * Feed everything into the reader if something doesn't have RSS try to came up with something that will convert that into RSS, e.g. Web Scrapping or Puppeteer (the last one would probably be required to Facebook).