HN Explorer

Ask HN: could HN threads be upvoted using machine learning?

1 points by highCs ↗ HN
Could HN threads be upvoted using machine learning based on the comments (read/write behaviors)?

EDIT: in a way that would satisfy the current HN audience of course

4 comments

[ 5.2 ms ] story [ 13.6 ms ] thread
tlb
By the time threads start getting multiple comments, they've gotten enough exposure from humans to upvote them to the front page if they deserve it.

More interesting is to scan new submissions, download the article and run a model to predict whether it's interesting enough to upvote. You can start by scanning old articles and trying to predict votes.

A major challenge is detecting dupes, follow-on articles and blogspam where the text looks relevant, but the topic has already been covered.

highCs
> More interesting is to scan new submissions, download the article and run a model to predict whether it's interesting enough to upvote. You can start by scanning old articles and trying to predict votes.

While I see why you see that as interesting, I think that submissions should not be filtered by a computer based on content for quasi-ethic reasons.

tlb
Do you believe any automatic content-based ranking is unethical? Even filtering exact duplicates? Or just some? If so, can you elaborate?