Previous discussion from when this was news:
Google's PageRank patent has expired (2019)
141–150 of 173 posts
Re: Google's PageRank patent has expired (2019)
#142I'm not sure how useful this might be to anyone anymore. Google still uses it for the base ranking. But, the results are then run through a variety of add-on ML pipelines, like Vince (authority/brand power), Panda (inbound link quality), Penguin (content quality), and many others that target other attributes (page layout, ad placement, etc). Then there's also more granular weightings for things like "power within a n…
Whatever you do, whichever ML model you develop, a query-independent ordering of all documents will always be necessary since all distributed IR systems will over-retrieve according to some form of term matching and then apply sophisticated scoring to extract the best documents. You can't score trillions of docs for every query. Google still uses PageRank, but at the risk of stating the obvious, the current PageRank…
> You can't score trillions of docs for every query.
That completely depends on how you model the queries. It can be a TB sized relation all the way into needing more bits than there are atoms on the universe.
Re: Google's PageRank patent has expired (2019)
#143Earlier quoted context omitted.
My search results were a lot better in 2006 when, I assume, they didn't have all these ML pipelines...
Your search results were a lot better in 2006 when there wasn't yet a huge population of sites cranking out SEOed garbage. Google is trying, they're just losing the arms race of detecting crap content vs generating it.
Why not crowd-source it?
All Google needs is an obvious, one-click "spam" button for logged-in users. Clicking should add a site to the user's spam blacklist (which they should be able to review).
They know about what users search for and how those searches overlap and separate users into groups (not to mention all the individual details they have access to). When sites are marked as spam by enough different types of users, those sites can then be manually reviewed and their content blacklisted by Google preventing the same or very similar sites with a different URL from appearing.
Unfortunately, Google makes a lot of money from these ad/spam sites, so they have a perverse incentive to keep allowing them.
Re: Google's PageRank patent has expired (2019)
#144Earlier quoted context omitted.
That's because the web was better in 2006. I think Google has made one change for the worse, though, which is strongly favoring more recent content. Increasingly, I think that change has been a big contributor to the decay of the web since.
Part of the problem is, I think, many searches strongly benefit from up to date content- programming tools, fashion, celebrities, things to do in X area, etc. It seems that Google has decided that most people want the most updated information when they look for something, which I don't think is entirely unreasonable. What I would love , however, is a way to turn that off for particular searches. Researching past even…
> What I would love, however, is a way to turn that off for particular searches
But you can! In search results, click Tools and switch the Any time dropdown to Custom range... and you can specify a date range in the past. (Apparently, the custom option is hidden in the mobile version?!) I'm not sure how precise and dependable it is but it seems at least partially useful when I search for historical events.
Re: Google's PageRank patent has expired (2019)
#145Earlier quoted context omitted.
> If you check my original comment I gave a simple scheme to attack HITS rank. Sybil attack is a little harder for Pagerank. OK, but how is it harder for PageRank? I can't really see any differences in the semantics of the two algorithms, so I'm not sure what kind of added vulnerability one or the other could have. > One could pose this as an adversarial game. Yeah, I appreciate that, that's what I was referring to a…
> OK, but how is it harder for PageRank? If you are familiar with the algorithms, which I assume you are, you can work it out. To make my page score high on the PageRank score I need to acquire links from high PageRank score pages. This is a lot harder because it depends on a) in-links and b) high PageRank pages. With Hits, its easy for one page to harvest a high Hub score. All that is needed is to outlink to known g…
> With Hits, its easy for one page to harvest a high Hub score. All that is needed is to outlink to known good pages (authority). Providing outlinks is trivial. Once so harvested, one can direct that flow to a designated page to give it a high Authority score.
Are you saying that you think HITS doesn't recursively score the quality of references by their own scores? That's not true. It does exactly what PageRank does in that respect: a page's score depends on the score of those which reference it, which in turn depends on... etc.
The 'hub' vs 'authority' distinction is interesting but not really relevant here: we're considering a page's 'authority' score, which depends on the 'hub' score of those who outlink to it, and at that point we're just doing PageRank [again, except performance-wise and arguably freshness-wise].
Like I said: the only non-trivial differences between them are implementation / performance-related, not semantic.
> Why not ? Formalize the strategy spaces of both the players and its a very valid game in the Game Theory sense. For the ranker you have to consider some functional space of functions over a graph.
Yes, again: possible to frame it as a formally valid problem if you really want to; still not an interesting one. We're only talking about this because you want to maintain that your earlier statement was true.
"You have to consider some functional space of functions over a graph" gives no detail (besides that, yes, you can model something–maybe documents, maybe people, who knows?–as a graph) and sounds like something written by a person with a gun to their head.
Or maybe I'm wrong and there's a fascinating problem which you just don't want to divulge to me.
Re: Google's PageRank patent has expired (2019)
#146I'm not sure how useful this might be to anyone anymore. Google still uses it for the base ranking. But, the results are then run through a variety of add-on ML pipelines, like Vince (authority/brand power), Panda (inbound link quality), Penguin (content quality), and many others that target other attributes (page layout, ad placement, etc). Then there's also more granular weightings for things like "power within a n…
My search results were a lot better in 2006 when, I assume, they didn't have all these ML pipelines...
Man I have such a different search experience than everyone on HN. Google is amazing for me. Can someone give me a search query they think gives "objectively" bad results and maybe some links they would expect to see who up in the top that aren't there? Or is it that you search for something and you don't find anything? Or is it just that people don't like that there are ads?
I genuinely don't understand.
Re: Google's PageRank patent has expired (2019)
#147Earlier quoted context omitted.
Part of the problem is, I think, many searches strongly benefit from up to date content- programming tools, fashion, celebrities, things to do in X area, etc. It seems that Google has decided that most people want the most updated information when they look for something, which I don't think is entirely unreasonable. What I would love , however, is a way to turn that off for particular searches. Researching past even…
You mention options in passing, but it is to me the root of the problem: Google hates giving up control. Control means ad revenue. So we could have options that would make search extremely efficient for most users, but that would presumably be very hard to monetize in comparison. So we have no options, and everyone gets mediocre to bad results. Since everyone I know in tech laments Google's decline into uselessness,…
> everyone gets mediocre to bad results
Everything I've heard from people I know at Google suggests otherwise. Most searches for most people ... work. I too struggle to have google work in specific research cases, and I would like more power-user toggles, but basic searches like "$celeberty_name photos" or "$my_kids_school calendar" or "pizza places near me" just sorta work.
Re: Google's PageRank patent has expired (2019)
#148Earlier quoted context omitted.
> My search results were a lot better in 2006 when, I assume, they didn't have all these ML pipelines... That's like an old person complaining that their body felt a lot better back in 2006 when, I assume, they didn't have to use their walker and glasses all the time...
Search is objectively worse today than it used to be, it's not just that "it's harder to use for old people", it's just worse.
I was trying to be clever with my metaphor, but since a lot of people seemed to miss my point, I'll spell it out.
An old person's glasses and walker don't make their body feel worse. They're responses to an underlying change, and in fact make them feel better than they would without them.
Similarly, I'd argue that the ML pipelines and complexities in Google search aren't why search results are worse today. Rather, the web has changed with more SEO spam, walled gardens, content in videos, and search has changed in that you try to find more kinds of information than ever before. It's the underlying changes that make the search seem worse, and all of Google's fancy algorithms are imperfect responses to that. Without them, I'd be surprised if Google's results weren't far worse than in 2006.
The comment I responded to:
> My search results were a lot better in 2006 when, I assume, they didn't have all these ML pipelines...
made it seem like maybe the ML pipelines were somehow causing the decline in quality, rather than simply an imperfect response to changes in the web since then.
Re: Google's PageRank patent has expired (2019)
#149Earlier quoted context omitted.
I feel the same way, I think one of the biggest startup opportunities now is to create a decent search engine. Google is absolute rubbish and a new player needs to step in and take away that market share that google is throwing away by means of a terrible user experience.
I strongly disagree that there's a startup opportunity in Google's relatively poor search performance. Obsessive programmers like myself really want a search engine that does better and helps us find the weird corners of the web. Most people do not care. They can type a question into Google and get an answer, and that's all they're looking for.
And we have add a variety of competing search-engines that still aren't as good. It isn't like the idea hasn't been tried. For example DuckDuckGo, or even Bing, but there have been half-dozen others too.
Re: Google's PageRank patent has expired (2019)
#150Earlier quoted context omitted.
> way better than what’s currently available Hand-curated walled garden sub-web that bans SEO spam with an iron hand.
wikipedia? Beyond their remit for some queries for sure, but they fit the mold.
A search engine that only searched sites that wikipedia links to might be a fairly decent source in fact.
If monetisable, it would turn gaming wikipedia into a whole new level of shitshow of course.