Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

PageRank isn't new at this point, has been publicly available for more than 10 years. If I were less of an idiot myself I'd found a fresh search engine. Other than Dear Bing (who doesn't support the "Verbatim" filter), there are no public independent full-web indexes left.

R.I.P.



What makes you think that Search still runs mainly guided by page rank?

If that were the case then search engines would differ only on how big their database is, but unless you care about the long tail you wouldn't be able to notice much of a difference.

I don't work in ranking, but considering how the web has changed since the early 2000s I'd guess that page rank's quality has gone down as spammy websites and walled gardens appeared.


PageRank is basically worthless at this point. Its baseline assumption, that keywords and in-links correlate with relevance, is now broken. The vast majority of stuff out there is SEO'd and link-farmed to death. If you want to find anything relevant with PageRank these days, you first need to find a way to filter out all of the spam. This is why search is so hard now.


> find a way to filter out all of the spam

It is the same spam sites I see again and again. A manual blacklist from one puny FTE would probably clean up the wast majority.


But that would involve hiring a human. Google's prime directive is "Never hire a human do a job well, when you can train AI to do the same job badly"


They could even crowd source this by allowing users to block domains from search results, which should affect its ranking long term.


As with so many things, they used to do this then inexplicably stopped.


You can use -inurl:example.com


Impossibly difficult and resource intensive? Sounds like we need a reset of some kind.


https://arxiv.org/abs/1710.05649

Google now has moved to model based ranking. In the current (search-hostile) web environment, PageRank or whatever simplistic algorithm doesn't work anymore.


Search is a really hard problem.

And as the web scales, it's an increasingly expensive problem. (Servers, storage, bandwidth, etc.)

I'd wager that it's easier to launch the next SpaceX than to launch the next Google.


Full-web isn't something google can do anymore either, for better and worse.


i've got my issues with brave but iirc they're actually doing an independent web index not just using bing (i think they still use them for images tho). it was really nice to hear somebody else is working on that even if i'm not using it rn.


I have issues with the Brave browser but I am a big fan of their search product. Just the fact that they are building their own index was enough for me to switch, but I've had better luck with finding useful search results than I previously had with DDG. YMMV, of course. Their forum search results section has been very useful, too.


Didn't know about Brave's Index. I just emailed them to find out how I can help with the bottom line of making human knowledge accessible for lesser-idiots. Thanks!




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: