|Google patent for PageRank|
Has anyone actually read the patent?
Google always says that their PageRank algorithm is patented.
Has anyone actually seen or read the patent though?
What does it say about how PageRank works?
I also wonder if all of the continuous changes and tweaks Google make to their algo really could end up invalidating their patent, since the beast they patented probably does not resemble the one they are using today, or the one they may be using tomorrow.
Therefore they couldn't really argue for intellectual property protection for their current algo, and should open source it or open it up for peer-review :-)
After all, even highly sensitive cryptographic algorithms like DES, AES, BlowFish etc etc are fully available for academic review.
The Google algo(s) should be robust enough to face up to this sort of open scrutiny.
Or maybe the reality is that PageRank algos have really just evolved into a mega-kludge, assembled ad-hoc with spit and string, with no real elegance or logic anymore.
This certainly appears to be the case, if my experience of Google's (degrading) search quality is any guide.
Not so sure about "their PageRank algorithm"...;)
|The first version of the PageRank technology was created while Larry and Sergey attended Stanford University, which owns a patent to PageRank. The PageRank patent expires in 2017. We hold a perpetual license to this patent. In October 2003, we extended our exclusivity period to this patent through 2011, at which point our license is non-exclusive |
Method for node ranking in a linked database [patft.uspto.gov]
as posted here in 2001: [webmasterworld.com...]
some more papers related to the creation of Google: [webmasterworld.com...]
Aha! That's great.
Thanks for posting the USPTO link - I'll now have a read of same in depth.
But interestingly, one of the first things I noticed was this:
"STATEMENT REGARDING GOVERNMENT SUPPORT
This invention was supported in part by the National Science Foundation grant number IRI-9411306-4. The Government has certain rights in the invention. "
So I guess this means, aside from Stanford U., Uncle Sam also has a stake (effectively) in Google?
I wonder how that pans out financially, given the IPO etc.
not too sure about the situation in the US but in Germany universities are owned by the government. And the rights for the patent have been extended till 2011 as stated above. So everything should be cleared imho.
Be careful in your decision to read the patent or not - it is not illegal to come up with something that might happen to resemble patent, but if you actually read the patent and then did it, then it could be interpreted as wilful infringement.
IMHO things like this should not be patentable. Oh wait, maybe its a good for the guy who gets patent, does not matter if it restricts everyone else as some very obvious things get patented and then enforced. Oh wait, maybe its good thing for whoever patents things, I am off to patent bubble sort and charge all those unis teaching it a hefty fee!
By our calculations the algo changes anything up to 9 times a month, the average being 5.
While this makes a mockery of those who post regarding their knowledge of how the algo works, based on something they noticed last week, or last month, I cannot see any evidence that this has any effect on the PageRank patent.
The important thing is that the Google algo has lots of different factors that play together to build the SERPs. PageRank is one of many, many factors, but should not be over-estimated. That's why I tend to say that there's always no corelation between PR and position in the SERPs.
Besides there being many factors in addition to PageRank to calculate the SERPs, there are also many factors that actually determine PageRank itself - as a good read of the patent will show.
Personally, I think Google is now assessing all those elements that can be be used to calculate PageRank more vigourously and that is the cause of the upheavals that have occurred over the last 12-18 months.
I would say that PR is still king - but not the value you see on the toolbar.
> I would say that PR is still king - but not the value you see on the toolbar.
I would say the opposite :)
|Be careful in your decision to read the patent or not - it is not illegal to come up with something that might happen to resemble patent, but if you actually read the patent and then did it, then it could be interpreted as wilful infringement. |
Independant inventions still infringe patents. Even if you invent something completly independant to the patent holder, you will still need a license from the patent holder. The whole point of the patent system is to get people to publish their inventions, in return for a limited period of exclusive control over that invention.
|Independant inventions still infringe patents |
Indeed, but infrigement based on lack of knowledge of all patents out there will create a lot less of a problem than wilful infringement.
I also think that knowing subject might incline you to think along similar lines as opposed to having to having a chance of developing substantially different work, thus avoiding infringement.
AFAIK this is the reason why, for example, creator of Linux does not make any patent searches.
|AFAIK this is the reason why, for example, creator of Linux does not make any patent searches. |
The software industry is built on the principle of Mutually Assured Destruction. Everyone holds patents to use to counter-attack anyone accusing them of patent infringement. ISTR IBM recently announced they would use their patent portfolia to crush anyone who used patents to attack Linux. The normal result is everyone cross-licensing their patents and forgetting it. Didn't Yahoo and Google recently do this over Adwords infringing on one of Yahoo's patents?
|Didn't Yahoo and Google recently do this over Adwords infringing on one of Yahoo's patents? |
Agreed about MAD. And this is exactly where the problem is - instead of cross-licensing (Google got no patents?!?) Yahoo tried to enforce its patent through courts and ultimately preferred to settle to get money quick (in form of shares).