news VS hnrss

I think so. I work with a group that was researching ways to improve the hacker news ranking algorithm. We concluded that there were "false positives" and "false negatives" -- stories often did or did not make the home page based more on timing and luck than the actual content of the story. Read more here: https://github.com/social-protocols/news

However, we concluded that the biggest problem is false negatives -- good stories that never made the home page -- because of submissions to /newest that simply don't get enough *attention*. There are too many stories and not enough eyeballs. As a result, there is not enough data to know if a story would do well if it were shown on the front page or not.

You can see this from the fact that there are many stories that are submitted multiple times, and 4 out of 5 submissions get only 1 or two no upvotes, and then the 5th makes the front page and gets hundreds. Sometimes it's due to timing factors -- the story has become relevant for some reason -- but often its just luck.

It's a short intro to raise awareness for the topic. Earlier, our research group "social protocols" published a new metric for Hacker News, but at that time, people didn't see the problem: https://github.com/social-protocols/news

Yesterday, we published our approach to improve the HN ranking algorithm: https://news.ycombinator.com/item?id=35183317

As a follow-up, we'd like to ask if the community is interested in official experiments with this or similar approaches. We'd suggest having a second, experimental front-page on HN, which applies the upoteRate metric to the HN formula, as described here: https://github.com/social-protocols/news#a-proposed-new-formula

To get a feeling, how stories with a high upvoteRate look like, we prepared a page, where all frontpage stories are sorted by upvoteRate: https://news.social-protocols.org/best-upvoterate.

What do you think?

Great to hear! We're still looking for strategies to cheaply retain more data. If you have any ideas, please jump in: https://github.com/social-protocols/news/issues/54

As a consequence, success on HN depends almost entirely on getting enough upvotes in the first hour or so to make the front page and get caught in this feedback loop. And getting these early upvotes is largely a matter of timing, luck, and moderator decisions. And so the best stories don't always make the front page, and the stories on the front page are not always the best.

Our proposed solution is to use upvoteRate instead of upvotes in the ranking formula. upvoteRate is an estimate of how much more or less likely users are to upvote a story compared to the average story, taking account how much attention the story as received, based on a history of the ranks and times at which it has been shown. You can read about how we calculate this metric in more detail here: https://github.com/social-protocols/news#readme

About 1.5 years ago, we published an article with this basic idea of counteracting the rank-upvotes feedback loop by using attention as negative feedback. We received very valuable input from the HN community (https://news.ycombinator.com/item?id=28391659). Quality News has been created based largely on this feedback.

Currently, Quality News shows the upvoteRate metric for live Hacker News data, as well as charts of the rank and upvote history of each story. We have not yet implemented an alternative ranking algorithm, because we don't have access to data on flags and moderator actions, which are a major component of the HN ranking score.

We'd love to see the Hacker News team experiment with the new formula, perhaps on an alternative front page. This will allow the community to evaluate whether the new ranking formula is an improvement over the current one.

We look forward discussing our approach with you!

Links:

Site: https://news.social-protocols.org/

Readme: https://github.com/social-protocols/news#readme

Previous Blog Post: https://felx.me/2021/08/29/improving-the-hacker-news-ranking...

Previous Discussion: https://news.ycombinator.com/item?id=28391659

That’s interesting.

I have predictive models that can predict if a headline (w/o the rest of the article and not considering the URL) will (a) get more than 10 votes and (b) if it does get more than 10 votes will the votes/comments ratio be more than 2 (which is roughly average)

The first model gets a ROC-AUC (see https://scikit-learn.org/stable/modules/generated/sklearn.me...) in the low 60’s (not good, the second model gets in the low 70’s (actually pretty good though it is a heat seeking missile for clickbait headlines) and my latest content-based recommender for RSS items gets almost 80. (I saw a paper that one system at TikTok gets about 85)

To do all that you need about 10,000 headlines and don’t get a lot of benefit from having more than 100,000. The ceilings on performance have more to do with the nature of the problem rather than my models: the same article can get submitted twice and get 0 votes one time and 200 the other time so it can never be as accurate as “is this an article about galactic astronomy?”

I had it ingest the HN comments firehose and found the amount of articles was overwhelming, my YOShInOn RSS reader now ingests the “best comments” from

https://hnrss.github.io/

together with 110 other feeds and actually I like the comments it picks out a lot. Now that the system is adding about 3000 items per day it might be able to handle a big feed like the comments firehose since now those comments are diluted with so many quality articles. For a problem like that you might want a two-score system with: (i) is it relevant? (something I like) and (ii) is it popular? (like Google’s PageRank)

I think you could make a model that compares comments in the best comments feed with other comments. I have tried formulating the problems above as regression problems where I try to predict the actual score and it does not work well because of the uncertainty problem but formulated as a classification problem for a score over a threshold it is easy to make a well-calibrated model that tells you “this article has a 20% chance of frontpaging” which is about the best anyone can do.

Subscribe via rss, so you can scratch the curiosity itch and each the FOMO, without coming to the site all the time and looking over the same things 20 times?

https://hnrss.github.io/

https://news.ycombinator.com/item?id=9491978

and this https://hnrss.github.io/

ps i’m ok with some % of false positives, but hopefully a sprinkle of OpenAI could keep that magically low?

thanks

It sounds interesting but I use https://hnrss.github.io/

Unless it had most of the features of hnrss.org I would not be able to use it.

Perhaps you could pivot your approach and submit a PR to hnrss for the feature?

Looks like there's an unmerged PR on the third-party hnrss project that would add this: https://github.com/hnrss/hnrss/pull/84

Check out below link to get a more customized, topic wise rss feeds.

https://hnrss.github.io/