Forgot your password?
typodupeerror
Patents The Internet

Google Patents Search Algorithm 367

Posted by michael
from the google-this dept.
blastedtokyo writes "Google gets the first web search patent. According to this News.com.com article, Google was able to patent how they crawl and rank web pages. They claim "an improved search engine that refines a document's relevance score based on interconnectivity of the document within a set of relevant documents.""
This discussion has been archived. No new comments can be posted.

Google Patents Search Algorithm

Comments Filter:
  • Good for them... (Score:5, Interesting)

    by DCowern (182668) on Thursday February 27, 2003 @09:50AM (#5395602) Homepage

    They thought of a way to improve upon an existing invention. They were the first to do it. They want to make money from their idea. It's only logical for them to seek a patent. I guess congratulations are in order!

  • watch out (Score:0, Interesting)

    by I Want GNU! (556631) on Thursday February 27, 2003 @09:52AM (#5395617) Homepage
    Here comes corporate Google!

    Whatever happened to the morals of Google "don't be evil"?

    I hope they aren't planning on trying to enforce this patent.

    As for the trademark, they actually expect to be able to restrict how people speak? They should be flattered that people use their name so often, it gives them publiclity. And generally people mean "to search with Google" when using the word. I know, I know, screwed up trademark law makes someone aggressively pursue a trademark or give it up, but still...
  • by Anonymous Coward on Thursday February 27, 2003 @09:52AM (#5395633)
    Does this mean that with their algorithm now publicly available, we're going to find more "googlebuster" sites finding ways to improve their rankings?
  • by ManUMan (571203) on Thursday February 27, 2003 @09:53AM (#5395636)
    I think we miss the point many times about patents. While it is certianly the case that many people use patents to protect "inventions" that have been around for a while, I think that patents should be granted for truly inovative discoveries.

    Perhaps the problem is that it is hard to decide when someone is actually seeking to protect a real discovery instead of using the law to pad their pocketbook.

    --JS
  • by the_burton (147439) <the_burton@hotmail.com> on Thursday February 27, 2003 @09:54AM (#5395649) Homepage
    I think that as time goes on, a companies ability to operate without the necessary business operations like patents will diminish. I guess what it comes down to is that if they want to stay at the top, then they have to have patents to protect their IP. Does it mean that some of google's shiny armour will be tarnished? Yes it does, especially in the eyes of all the geeks out there who see patents as the Great Evil. However, the company will remain in business for quite some time, allowing them to keep operating business as usual. So far, business as usual is good enough for me.
  • by terrencefw (605681) <slashdot@@@jamesholden...net> on Thursday February 27, 2003 @10:00AM (#5395725) Homepage
    I thought Google's searching/ranking tehcnology was a closely-guarded trade secret, to make sure that people weren't able to engineer their rankings sucessfully.

    Now that they've patented their technology, surely that means that it's open to public scrutiny and therefore abuse as people exploit it's shortcomings.

  • Re:Good for them... (Score:0, Interesting)

    by MartinG (52587) on Thursday February 27, 2003 @10:02AM (#5395744) Homepage Journal
    Actually, I don't believe they did invent it. They have stated themselves that they accidentally found their useful search algorythm while trying to devise something else (a system for rating pages IIRC).

    So, personally I would say they discovered it rather than invented.

  • by Anonymous Coward on Thursday February 27, 2003 @10:02AM (#5395749)
    Do you realise that the Google search you link to, shows your comment as the top result? Its a Google loop!
  • Prior Art? (Score:4, Interesting)

    by Greyfox (87712) on Thursday February 27, 2003 @10:10AM (#5395807) Homepage Journal
    Wouldn't google be... wait... nevermind.

    There's a reason I only ever use their search engine now. Well two reasons. One is that about half the time I run searches there, what I'm looking for is the first thing on the list. The second is they are very not obnoxious about their advertising. And I've probably clicked through more google ads than any other banner ads on the net. That's right, I'm much more likely to follow information that looks like it pertains to what I'm looking for right now over some obnoxious Javascript ad (Which usually make me turn Javascript off and reload.)

    Very similar to Google's method, I've seen CNN and USA Today run ads disguised as news stories in their tech sections. Unlike google, which clearly marks the ads, CNN and USA Today are simply compromising their journalistic integrity. As if those two words have been put together in a single sentence since Cronkite left the industry.

    Where was I? Oh yes. In principle, software patents offend me. Well... and most of the rest of the slashdot population apparently. Being able to patent something that doesn't have a physical presence (Be it programs or math or business processes) is counter-productive. Especially since the patent office seems to rubber stamp every application that hits their desk. Hey. If you don't like it, write a civil nastygram to your congresscritter. Do NOT use the word "Fuck." That tends to turn them off. And in extreme case, get you visits from very grumpy people who seem to have something against doors. We're starting to see some technologically clueful folks in office, so the more people who write, the higher the chance that someone in the know might get the message.

  • by expro (597113) on Thursday February 27, 2003 @10:14AM (#5395843)

    So, the bright side of this patent is that perhaps it will keep others from focusing on Google's obsession -- the reference popularity contest. But like any patent, it is subject to abuse, not that we know at all how Google intends to enforce it.


    I have requested improvements to Google's algorithms for years to make it more possible to search for a specific thing, rather than just a popular thing, but they don't have engineers, apparently, who understand these basic needs.


    AltaVista lets you wildcard, search for one word NEAR another word, use common words as part of a phrase, and construct a variety of very useful filters that are impossible with Google's popularity engine.


    AltaVista used to be the best out there, but compromised their own usefulness. If AV indexed more pages and had not dropped their usenet coverage, it would still be the most useful engine by far to an advanced searcher -- one looking for very specific things. I still go there often. Just because the masses use Google does not make it quality or best for advanced users. They have stagnated for years now. The masses use a lot of things produced by monopolists who are no longer required to innovate or even improve to the level of the competition.


  • Re:Good for them... (Score:5, Interesting)

    by PygmyTrojan (605138) on Thursday February 27, 2003 @10:20AM (#5395890)
    Mod it down or reply, that is the question...

    Hopefully you've read what you wrote by now and realized how stupid of a comment that was. Gravity was a discovery becuase it already existed. Algorithms are invented. I don't know about you, but I haven't seen any algorithms on the side of the road when I drive to work.

    Besides, most inventions were found by accident, that doesn't make them discoveries.

  • by Anonymous Coward on Thursday February 27, 2003 @10:25AM (#5395925)
    Google is a pretty good dumb word matcher but it falls flat on its face with verb tenses. For example: "run", "running" and "ran" are different words for Google. The reason is simple - Google makes no attempt to understand language - english or otherwise. Why is this? Because it is difficult, ambiguous and computationally expensive. The ultimate web search tool has to definitely improve in this area.
  • Re:Mis-title (Score:5, Interesting)

    by johnnyb (4816) <jonathan@bartlettpublishing.com> on Thursday February 27, 2003 @10:27AM (#5395945) Homepage
    What I found particularly cool about their algorithm was that they can return results for pages that google has not read. If there is a link to a page google has not read on a page that google is reading, it can still return results to the unread page based on the context of the link, and the popularity of the link on other pages. Really nifty stuff.
  • by Sydney Weidman (187981) on Thursday February 27, 2003 @10:29AM (#5395964) Homepage
    Right, that's a very interesting and important point. You can kind of imagine the sequence of events unfolding like this:
    • Public scrutiny forces a redisign of the algorithm
    • The redesigned algorithm cannot be covered by the original patent
    • Either Google applies for a new patent covering the improved algorithm, or they let the algorithm become public domain, since it can no longer be a trade secret
    • which means the money spent on the development and the patenting process (which might be on the order of 2 million USD) is wasted
    Doesn't this indicate that software and business method patents do more to waste resources than generate innovation? You're hamstrung -- unable to improve your new moustrap because if you do, it won't be protected anymore. Or you'll have to go through the whole expensive process of applying for a patent all over again. This doesn't sound like a very strong incentive for R & D. It sounds more like a pain in the ass.
  • by Lxy (80823) on Thursday February 27, 2003 @10:36AM (#5396023) Journal
    Dude, that is the WRONG way to look at this.

    NOT ALL PATENTS ARE BAD. If Jeff Bezos patented a specific algorithm for searching Amazon.com, something radical that no one has ever conceived, he should be awarded a patent.

    What you're thinking is this:

    Jeff Bezos patents a "method of picking your nose by sticking one finger up it and turning 90 degreees", we get mad at him and coplain that the patent system is broken.

    Google patents "a method of typing text into a box and clicking submit", you ignore it JUST because they're Google. Wrong attitude.

    This is a GOOD PATENT. Google built an algorithm, an algorithm so complex that no one has been able to figure out how it's done. No one has been able to copy it, no one can imitate it. Long, hard hours by the Google team has paid off. They produced a patentable product, and they deserve a patent on this product. They're not stifling innovation, they're protecting their work.

    The basics of patents:

    patenting a widget is a GOOD PATENT
    patenting a method of using the widget IS NOT
  • by Quetza (5618) on Thursday February 27, 2003 @10:43AM (#5396089)
    There seems to be a lack of understanding about the original purpose of the patent system. The the distant past, knowledge was transferred from artisan to apprentice and through guilds. Back then, as now, people were very protective of their intellectual property, as it was their livelihood, so it would not be stored anywhere. If the person were to die without passing on the knowledge, it would be lost forever (like Damascus steel).
    To try and stop knowledge from being lost, governments introduced a patent system (first patent recorded in 1449 [derwent.com]) so that the creator of the knowledge would still get a fair financial reward for the item.

    IMHO there are 2 problems with the existing patent system implementations.
    1) As the technology becomes more complicated, those who verify patents are not skilled enough to accurately judge their validity.
    2) The time limit of patents is too inflexible. Many technology patents should have valid lengths of 5-10 years.

  • Re:Good for them... (Score:5, Interesting)

    by Qzukk (229616) on Thursday February 27, 2003 @10:52AM (#5396184) Journal
    The problem with the patent system today (in my view) isn't whether or not software should be patented, it is the complete crap that comes out of the PTO. They have no incentive to do things right, no disincentive against doing things wrong, so they just grant patents on crap.

    What needs to be done is first: fix the backlog so that it no longer takes years to grant a patent. Open new patent courts, hire more people, charge more for the patents, whatever it takes.

    Second: The USPTO needs to be responsible for its output. If a patent is overturned as invalid, court fees and the original cost of filing the patent should come out of its pocket.

    Third: Mandatory implementation. For a patent to be valid it must be implemented by someone, somewhere. If the holder of the patent is unable to implement the patent, it must accept at least one request to license the patent for implementation under generally accepted RAND practices. This prevents people from patenting something with no intention of ever producing that invention, for the purpose of preventing that invention from being produced.

    Fourth: establish a class of patents specifically for software. Things not in this class of patents cannot be used to claim software infringes. Turnaround must be quick, and part of the patent process should be "does someone else sell this product already" which should be relatively easy, compared to looking through millions of patents.

    Software patents will expire after two years, renewable once (at the price of the original patent) for another year. This better matches the reality of software development. It provides people a head start without granting them an essentially perpetual monopoly.
  • by pheph (234655) on Thursday February 27, 2003 @10:53AM (#5396199) Homepage
    Google is really breaking out the lawyers lately... They are currently threatening a non-profit (no advertising, no pop-ups, not for sale) site, which clearly was not google.com or associated with google.com (gewgle.com [gewgle.com]) for trademark infringement, bad faith under UDRP, and more. We have gone back and forth a few times, check out gewgle.com's legal section [gewgle.com] or just gewgle.com [gewgle.com]
  • by NoOneInParticular (221808) on Thursday February 27, 2003 @11:55AM (#5396917)
    you wrote:

    If someone can figure out how to achieve the same result with a diffrent formula, more power to 'em!

    okay, page rank is (matlab)

    r = ones(1,length(W)) * W

    Where W is the link matrix of the WWW. NewRank is:

    r = ones(1,length(W)) * (Q * W)

    Where Q is a diagonal matrix representing the results of a query. Now would my innovation

    r = (W' * ones(size(W),1))'

    Be considered the same result as page rank with a different formula? .

    Or, mathematically speaking, wouldn't achieving the same result with a different formula, prove that the formula is identical to the original, only using a different notation? Hence, is it mathematically possible to patent the formula and not the fragrance?

    This is one of the many reasons that algorithms should not be patentable.

  • by Anonymous Coward on Thursday February 27, 2003 @12:13PM (#5397139)
    My initial reaction was "it's ok, they are patenting a truly new idea". Then I read the patent. The patent specifies an algorithm. In one sense that's good, the patent is not wide open, it's not like they are patenting "web searches" or "database indexing", they are patenting a very specific way for building indices and extracting information out of the database. On the other hand, this is bad. They are patenting an /algorithm/. They are not patenting an specific rendition of an idea, they are patenting the idea itself. This stops research. You can't do basic research that improves on the idea, because you run the risk of your modification not being /different enough/ (there's plenty of examples of algorithms where you add a little detail to make them work better on specific cases, or add details to make them more general, or variations like that). If you modify the algorithm like this and you publish your results, you /would be/ infringing on Google Inc.'s patent.

    This is the reason why patents on algorithms are bad. Unlike patents on physical devices (think the common mouse trap, a cage with bait inside vs. a spring trap), with algorithms it's extremely difficult to draw the line. What would happen if someone tried to patent a mouse trap. The idea of baiting a mouse and then [???] and thereby trapping the mouse? It would be rejected. You can't just patent the intention of trapping a mouse. You have to be more specific than that. With algorithms, it might seem that you are being specific enough, when in fact you aren't.

    Interestingly enough, Google's patent applies only to searches over the network (that, I really can't read the languaje used in patent claims).
  • by Minna Kirai (624281) on Thursday February 27, 2003 @02:20PM (#5398795)
    through diligent study you may be able to determine whether or not you are in violation of some particular software patent

    Moreover, if you do attempt to keep aware of software patents, then in the event you are found guilty of violating one, the punishment will be worse, as it was willful, knowing infringement.
  • Re:Why not? (Score:3, Interesting)

    by Fulcrum of Evil (560260) on Thursday February 27, 2003 @04:11PM (#5400039)

    It is relevant because the idea that strong IP necessarily reduces innovation is clearly false.

    A position contradicted by a recent slashdot article. The article in question is far more eloquent than I am, so I will just mention that most of the innovation in this country happened in spite of strong IP. Strong IP protection protects the established power structure - not a recipe for innovation.

    Why is it that China and India produce so little of note, whether you're talking about scientific research or technical/product innovation, despite having tons of educated people and little to no concept of IP.

    Well, China is a brutal dictatorship where you can be disappeared for thinking differently, and India, while probably better in this respect, still defines social standing in part by whether you get enough to eat. Never mind that India is fairly protectionist with its trade policies.

    Why is it that the USSR did so little when no one could "lock ideas up" for themselves?

    Well, the soviets practiced the same sort of groupthink that the chinese do - they chose scientific theories based on the political position of the scientist who proposed them instead of proven results, they centralized their entire economy, so any sort of innovation was difficult to impossible, and, yes, there was no incentive for innovation. It didn't matter that IP wasn't there - if you had a patent on some process, how were you to benefit when the state owned everything and they'd just take it anyways?

There are never any bugs you haven't found yet.

Working...