Hacker News Reader

Comment by RiderOfGiraffes | Hacker News Reader

My problem is that the items I'm most interested in never really get the upvotes to pass the thresholds, so I'm looking to train a Bayesian filter to match my interests, and rank items based on that.

I'm sure I'm not the only one.

View on HN

But isn't that the point of a social news website that the community (whose members should of course have similar preferences as you have) does the filtering for you by upvoting interesting articles. Otherwise I could run such a filter over a collections of rss feed aggregators, couldn't I? By using a social news website, users IMHO implicitely claim that community-based filtering works better than any algorithm could, don't they?

I hit the new once or twice a day, but sometimes it s just too much.

Plus I don't feel that welcomed when the only enhancements to the site I see are taking away permissions (I'm at 460 karma. I can't downvote anymore)

Perhaps you miss my point. I said:

  > ... the items I'm most interested in never
  > really get the upvotes to pass the thresholds,

I've found dozens of items that I'm interested in and which never got more than one or two upvotes. They never made it to the front page. Often they've been resubmitted, the timing has been right, or the title has been changed, and suddenly it's got tens, on some cases hundred of upvotes.

There are many, many items that slip through, unnoticed, unloved, that should get upvotes, but due to poor timing or poor titles don't.

I want a system that helps me catch them.

Great idea. Let me know if you do that. One could extend it so that it could find users with similar interests.

You would probably need a 'down vote' button to say "I'm not interested in this"

Reply on news.ycombinator.com