7 ms·

Because ranking is very difficult

metadat 4y ago
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.
- summerlight 4y ago
  https://arxiv.org/abs/1710.05649 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.
- monetus 4y ago
  Full-web isn't something google can do anymore either, for better and worse.
- dietr1ch 4y ago
  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.
- collegeburner 4y ago
  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.
  - metadat 4y ago
    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!
  - godshatter 4y ago
    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.
- ijidak 4y ago
  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.
- chongli 4y ago
  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.
  - metadat 4y ago
    Impossibly difficult and resource intensive? Sounds like we need a reset of some kind.
  - rightbyte 4y ago
    > 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.
    - graftak 4y ago
      They could even crowd source this by allowing users to block domains from search results, which should affect its ranking long term.
      - jpindar 4y ago
        As with so many things, they used to do this then inexplicably stopped.
        anticensor 4y ago
        You can use -inurl:example.com
    - nikanj 4y ago
      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"

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.

summerlight 4y ago
https://arxiv.org/abs/1710.05649 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.
monetus 4y ago
Full-web isn't something google can do anymore either, for better and worse.
dietr1ch 4y ago
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.
collegeburner 4y ago
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.
- metadat 4y ago
  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!
- godshatter 4y ago
  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.
ijidak 4y ago
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.
chongli 4y ago
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.
- metadat 4y ago
  Impossibly difficult and resource intensive? Sounds like we need a reset of some kind.
- rightbyte 4y ago
  > 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.
  - graftak 4y ago
    They could even crowd source this by allowing users to block domains from search results, which should affect its ranking long term.
    - jpindar 4y ago
      As with so many things, they used to do this then inexplicably stopped.
      - anticensor 4y ago
        You can use -inurl:example.com
  - nikanj 4y ago
    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"