Site Reference Forums

Originally I wrote ‘ Phrase Based Optimization ’ – this is the more technical analysis and may not be for everyone.

Next was ‘ Phrase Based Indexing and Retrieval ’ – which breaks down what it means to us in some simpler terms.

k, so im gonna study these two, print em, and remember them to the word.

Strangely, I stumbled across some evidence of PaIR in the recent demolition of GoogleBombs – Our pal here Jaan ( IncredibleHelp ) posted some info as wellwhich led me to Stoney (DeGetyr) and we’re following up my little find…

Would this be useful in context to the previous two? and in such be worthwhile memorising with the above or would they be classified as still up for debate?

Conspiracy, i like conspiracies....
eg.... why is it that a man must go through the effort of lifting the toilet seat and putting it down, where does meeting half way come in? surely the effort taken to lift it is equal to (if not less than) the effort to drop it, so why to woman get to complain about it being up while men are frowned upon when complaining about it being down? Now thats a conspiracy

Ryan_steyn wrote:

Conspiracy, i like conspiracies....
eg.... why is it that a man must go through the effort of lifting the toilet seat and putting it down, where does meeting half way come in? surely the effort taken to lift it is equal to (if not less than) the effort to drop it, so why to woman get to complain about it being up while men are frowned upon when complaining about it being down? Now thats a conspiracy

Just learn to sit - it'll save you having to clean up when you get much older!!! 

jmac wrote:

It makes a lot of sense when combined with the Scottish Web Design experiment. However, it does muddy the waters for the beginners like me.  Whilst writing 'less than stuffed' content is what we've learned, what I'm now learning is that we need to broaden the scope to related items whilst maintaining the main topic. I think.... or did I get it wrong again.

lol

Think of it this way.....

Over a given document or set of documents ( web pages and web sites) there will be a statistical average for usage of related phrases to a given term.

Now when it indexes your page it compares it to this statistical average of related phrase usage.
For example

President of the United States

it would also expect to find things such as

George Bush
Bill Clinton
the White House
Democratic Party

and so on..

From this AVERAGE ( of topical documents in the Index) a threshold/baseline is set.

YOUR document (web page) is weighted and scored against the basline. It is also compared to the thresholds ( too high or low can be flagged as Spam)

APPARENTLY - the average site has between 8-20 'good phrases' whereas a Spam document(s) can have between 100-1000 'good phrases'

Is that getting U somewhere?

Last edited by TheGypsy (2007-02-06 08:42:29)

Ryan_steyn wrote:

APPARENTLY - the average site has between 8-20 'good phrases' whereas a Spam document(s) can have between 100-1000 'good phrases'

Now is that per page or website? I can see that confusion getting messy

BOTH.

IF YOU READ the articles I posted ( kinda why I did eh)... it looks at documents as well as sets of documents (remember there are PDFs and other documents on hte web than Web Pages).

but YES, it looks at the aggregate as well as the individual documents

I understood that LSI influenced the SERP of my site, while PaIR can identify me as a spammer, right?

This brings up one question for me: how much filtering could a visitor take?

Last edited by nybc (2007-02-08 02:50:41)

OK, so my brain is back well enough to take on investigating your links, Dave. Now that I have, got a few questions --

1.) SEO was done (you might want to see my last question, since your answer will indicates if this is even a relevent question anymore) by setting up our sites to fit what the SEs consider important – good domain names, page names, KW phrase optimizing, meta tags, Internal links, BLs verses RL, headers and alt images – as well as, setting it up to be user friendly. On the grand scheme of things, is there a relative hierarchy for all these elements, or are all these things too important to decide which is/are the most important to SE? If there is a relative and necessary hierarchy, what is it?

2.) What’s anchor phrase and/or anchor text?

3.) Gypsy wrote (in his second article mentioned):

“What’s a ‘Good Phrase’? The classification for possible phrases as either a good phrase or a bad phrase is when the possible phrase; ‘appears in a minimum number of documents, and appear a minimum number of instances in the document collection’. What that number is, we don’t know. Those are the ‘dials’ the Search Gods themselves only have access to. It is almost looking at a Phrase Density over the aggregate of documents (the web site). Also, a BAD phrase is not one with dirty words; it is simply a phrase with too low a frequency count to make the ‘good’ list.”

(I might be panicking prematurely, but, well, can’t help it!) Ah, man, this stinks! Websites generally are about some central point. By the very nature of content, or, at least, relevant content, some phrases are going to come up repeatedly! If they’re terms that produce no synonyms from a thesaurus, apparently we aren’t supposed to mention the main subject matter?! All this time, I’ve been trying to optimize to filter my sites away from the commonplace merchants. I apparently can’t compete with them, and won’t even show up without them. Do we now disregard all the stuff we learned about optimizing – kill as many of the KW phrases as possible, particularly in alt tags, headers, domain names, and even in the content? My site (which is about to be relocated to a new domain name at another web host) has 10 pages and a link, therefore between the title on each page and the pages’ menu, I’ve already topped out to the 8-20 “good phrase.” Add to that my main KW phrase, both in content and in image names, now easily goes well past 100! Am I missing something, or did Google just create an entirely different IR (Indexing and Retrieval) formula?!

Well hey folks... seems they FINALLY started thinking about PaIR over at WebMaster World.... the first thread on it just peek out.. yesterday..

So, how's it feel to be ' ahead of the curve'??

He he

Well, the main problem folks are having – is the ‘how-where’- . It is NOT replacing the existing infrastructure – it is in concert with existing algorithmic IR methods. Certainly, over time, as the dials are turned in PaIRs favor, the other IR methods could be devalued.

Almost like the ‘onion’ analogy – in reverse – adding layers.

DMX wrote:

Odd coincidence that google and others would be refining the way documents are sorted or indexed for retrieval using roughly the same information they would to determine if it was duplicate content elsewhere, making a number of these threads a different facet of the same  mechanism..

How clever. We're pretty well being "forced" to write original, interesting and unique content. Hate that.

PaIR is a very comprehensive methodology which can deal with a great many areas of the IR process including (covered in the Patents);

The documents deal with;
Indexing
Weighting Ranking
Duplicate content
Spam detection and weighting (yes - not all spam is bad?)
Back Links/Link Profiles
Personalized Search
And much more.

As far as being ‘forced to write original content’ – I fishing hope so. One of my FAVORITE parts of all this is I can see it being a more effective method than is currently in place for searchers, webmasters and SEOs alike. IMO tho….

How clever. We're pretty well being "forced" to write original, interesting and unique content. Hate that.

DMX, actually, might not be as hard as you think. I graduated college almost 30 years ago with a degree in communications. With that (and needing marketing for my own business), I tend to observe marketing trends. One trend that’s been going downhill quickly over the last 30 years is advertisement. Back in the olden days, when I was young, commercials and print ads actually told benefits of the product. Now, if they even mention what product they’re selling, it’s just shown with a lot of glitz and glamour, or tough looking jocks performing with the “wonderful” catch all phrase, that just almost forces me to leap out of my chair, “Just Do It”

It’s not just sneaker companies either. When was the last time you were told some wonderful benefit to a mid-size car (“everybody is buying it” does not qualify as a benefit), or, yeesh, even what the heck that perfume smells like? (OK, I’ll buy a small bottle of White Diamonds, if it comes with that pair of earrings Liz Taylor is giving away for luck!) Want to give “original” content to sell your products online? Just find out “original” content used 40-50 years ago! Good solid copy that actually gives consumers’ benefits! All you need to do is update it, maybe.