Domain Registration

Google Search Algorithm Patent Application Creates Spring Buzz!

Google practical for a patent on their level algorithm as of 15 months ago on December 31, 2003 and that application was announceed on stalk 31st at the US Patent work. It got the argueion forums vivacious this weekend. Even still I had substantial work to do and was behind on a foretell, I couldn’t resist the temptation to read the very long 14,000 word, 45 page application and see what it could mean to the impulsive world of search.

So I tripped on over to the the US Patent & brand work (USPTO) and happening rendition the record United States Patent Application: 0050071741 seems to be Google employing for a patent on their search algorithm. There seems to be no situation to PageRank here, but it seems to be PageRank redefined with a few variations to boundary network spamming and moderate stale fallout, along with numerous innovative rudiments not previously thinked.

They argue network spamming boundaryations extensively, which would be a pleasing relief as involving neurosis is rampant and I’d like to see an end to it. greatly of historical records related to pages seems to be a bit arduous because it would begin to boundary the perceived worth of a page save it becomes wildly trendy over time. better is better seems to be a permanent theme of this algorithm as described generically in book of their application.

An odd addition to the historical level argueion is amazingly - the “Advertising travel” for a particular record! They will rank a location based on the presentr choosing to present on a particular location. If Amazon desires to present on your location, then Google will rank you senior!

That’s good, I deduction, if you have a location that attracts decidedly rated advertising, and don’t rely on impede promotion of your break harvest or those of suppliers to begin in your location advertising. Example: If I have a argueion forum on brunette, don’t I want to present my brunette harvest? Why would I wait ads from decidedly rated presentr Starbucks to rank senior at Google? What if I promote thousands of harvest and plainly impede promote and uppromote my own harvest locationwide? Odd things, level based on presentrs.

How does connect advertising make into that advertising aspect of the algorithm? Do they know you are advertising a book from Amazon as part of connect encode through your nonrest Amazon connect encode networks and do they grant tracking networks through connect management companies differently than the tracking URL’s of ad quota monsters like DoubleClick and discuss senior level advantageing the big boys of advertising above connect tracking firms?

Also seems to call into query their own Adsense ads and how that makes into this algorithm! Do the Adsense ads along my blog border expansion more level mark because it is from a monster advertising group - Google - or is it downgraded because I’m not a “Premium” presentr quota over 20 million content page views? Aexpansion, seems that reward for being large outweighs weight in this formula. Or does it? How do they worth approach advertising in the formula? Adbrite? slighter ad networks versus large advertising aggregators?

They extensively argue historical records related to levels over time, looking at seasonality, trendyity during spikes in travel due to reports coverage of a particular focuss and changes in level related to those stuff. The historical records related to level over time are attracting while they submit to network spamming, weight, and focusality when they say:

“As a expand evaluate to differentiate a record related to a focusal phenomenon from a spam record, search engine may think mentions of the record in reports articles, argueion groups, etc. on the idea that spam records will not be mentioned, for example, in the reports. Any or a combination of these techniques may be worn to restrain spamming challenges.”

They’ve added another attracting aspect in the algorithm of deidiomining worth of pages based on “customer maintained/generated records” (patent model 113) read that “bookmarks” and “favorites lists” built into your browser. Is this one of the reasons that Google freshly hired Ben Goodger, the advantage upgradeer of Firefox?

probing into my favorites and cookies on my structure seems like a bit more than I want Google burden on MY structure. It strains the boundarys of privacy as well. We can rest locations from quota us cookies, but can’t rest who reads them? Ouch!

foster, they situation customer’s browser hoard archive as a style of deidiomining worth of a location. “For example, the “temp” or hoard archive associated with customers could be checked by search engine to isolate whether there is an multiply or lessen in a record being added over time. equally, cookies associated with a particular record might be checked by search engine to establish whether there is an upward or sliding trend in attract in the record.” Apparently they can see this data, but I’d like them to halt out of my hoard and cookies too!

It begins to employ expand penalties to new locations by custody them poorly ranked for even longer intervals and applies an apparently new model to algorithms not seen or (at slightest argueed freely) of long idiom foothold of sphere names and historical records related to IP address and hosting group! Here’s the slice about that endurance of sphere registration to level:

“[0099] Certain signals may be worn to distinguish between illegitimate and legitimate spheres. For example, spheres can be renewed up to a interval of 10 days. dear (legitimate) spheres are regularly rewarded for numerous days in progress, while entry (illegitimate) spheres seldom are worn for more than a year. hence, the time when a sphere expires in the prospect can be worn as a make in predicting the legitimacy of a sphere and, hence, the records associated therewith.”

I’ll be extending the idiom of my sphere registrations ASAP! What a bonus to registrars if that aspect of level becomes as worthd as networking has been! each will get 10 year registrations if they want to rank well. The sphere name afidiomarket will also be distorted dramatically if this becomes as important as this aspect makes it begin to level. People will buy and promote spheres when disposing of them very than plainly charter them expire at the end of the registration interval, as most do now.

It begins they will be penalizing spheres “associated” with “illegitimate” spheres. Hopefully they have a style of deidiomining that it isn’t a competitor networking to your sphere from their “illegitimate” sphere! That suggests they will be able to eliminate “area Scrapers” that have been known to abrade search engine fallout of high level spheres and announceing those on “illegitimate spheres” which in realize drags down the level of those previously decidedly ranked spheres. How odd the search world is someperiod!

Altogether, it seems that adult content will undergo global because it hasn’t distorted, because nonentity new is networking to it and because it will escape networks over time. What if you are announceing a historical record that you can’t change or an authored model that is copyrighted? Does it lessen the worth of the datarmation? Hmmmm. I deduction networks would stay to multiply if the datarmation remainder worthy, so there is some protection in that. But adult location content may be undistorted because it is trendy, not because it is stale - that’s an odd seize-22.

The secure book make argueed in this patent application suggests that “[0118] exclusive terms, Bigrams, Phrases in secure copy ” are significant in deidiomining rank, because if real networks upgrade, they would alter when webmasters network to a record differently, some would use the URL and embed the network in that, others would use appealed book from the webmaster if it were a network appeal that successfully garnered a network and still others might plainly use Google’s own Blogger “Blog This” network which plainly takes the page name. (I routinely change network book generated by “Blog This” in my blog announces to underline the focus argueed and eliminate contract/publication names typically added early of the focus of the page.)

The US Patent workplace has a network to imagery counting illustrations and facts that are networked to the filing but they are absurdly large and don’t fit in the viewable framed skylight. This is absurdity. Do they mean to secrete it by making it unviewable?

I’ll challenge to announce a lesser form of imagery on my blog.

The finishing notable model seems to me to be the clickthrough records that Google sees to locations from their own search fallout. They will rank location senior that get significant clickthrough tariff from the Google SERP’s.

“Google may check the number of period that a record is chosen from a set of search fallout and/or the total of time one or more customers exhaust accessing the record. pursuit engine may then mark the record based, at slightest in part, on this datarmation.”

How will they know how long I exhaust accessing the record save they can check my actions AFTER I’ve left the Google SERP’s to stay the networked location? miracle what’s at work in that? Do they have some way of tracking our actions after we disappear their location? I marvel if this has something to do with the Google acquisition of brat travel statistics group last week.

Well, it’s back to work for now, but it will be attracting to see where this patent application is argueed in forums and SEO blogs over the imminent week.

Mike Banks Valentine is a pursuit Engine Optimization Specialist and blogs about the search world at: http://RealitySEO.com while working a small contract ecommerce tutorial at: http://weblocation101.com

Share This