66 replies to this topic

[Ruud]

Thank *you* for the effort and for the sharing. As you know from another post I'm not much of an SE-wathcer (yet) and this would have blown past me otherwise.

Not to say that your posts didn't make sense to me before but the emphasis you put on structural navigation, for example, makes *much* more sense now. It fits into the big picture.

The good thing is that it is still, or even more so then ever, about The Content. In the back of my mind I also have an idea about how this can negatively affect a page but haven't got it lined up yet. Basically it's simple though: make the 'mistake' of referencing the wrong material and the engine can be thrown off course.

I'm pretty sure that, yes indeed, Google is moving towards this. A search for "divx encoding will have doom9's divx-encoding.htm as result 1 even though the term appears only four times - but doom9 is all about video encoding. The word divx itself appears a lot more times in that page but a search for divx will rightly put divx.com at 1st place, followed by divx-digest.com - sites with are much more authoritive when it comes to just divx. And imo neither of these 2 results depends on links to them.

This is one of the most exciting 'discoveries' I've come across in a long time. In fact, it's almost as good as peanut butter !

Ruud

[torka]

Ruud, thank you as well -- your "moments of clarity" have helped advance my thinking on the subject, as well.

G and Vijay, thanks for posting the "suggested reading". The documents have been interesting reads and have got me thinking in all sorts of new and exciting ways about SEO. I'm actually looking forward to implementing my new thinking about this into my pages. Yes, it seems as though there will be more work (at least initially) to wrap my brain around a new paradigm, but in the end I think it's going to lead to richer and ultimately more visitor-useful content.

Keep talking! I'm still not sure I'm solid on all these topics. The more I learn, the better I like it...

--Torka

[copywriter]

All the info on this thread is wonderful! I've been investigating this stuff for about 3 weeks now. I've found several things out from reading and from actually talking with some folks I found that develop this type of technology.

1. Nouns are important. When the stripping occurs, nouns are specifically sought after and saved during the indexing.

2. Keywords will still play an important role in the semantic Web. Yes, many pages will be able to be found without any keywords at all. But, those that include keywords within natural language are likely to rank higher.

All of this has been said before, but I thought I'd add my 2 cents worth

Karon

[marin]

One great advantage of Latent Semantic Indexing is that it is a strictly mathematical approach, with no insight into the meaning of the documents or words it analyzes.
It can be used in conjunction with a regular keyword search, or in place of one, with good results.
Running the latent semantic model for retrieval is fairly cheap computationally, and increasing and decreasing the precision of the fit is very straightforward.
IMO Lsi explains stemming and why our singulars are well ranked.

Google is definitely interested in LSi: just try a search for "latent semantic" and look at adwords: WORK FOR GOOGLE . The adwords does not appear for "semantic" or other queries.

LSI tools here and Telecordia LSI search engine here

[powerofeyes]

Welcome Marin ,
Nice to see your posting here, Seems you have built some nice tools on "latent semantic indexing". Do you have any idea it might be used in google in any form other than singular words ranking, If yes please share with us,
How much you have researched on latent semantic indexing, do you work for colorado university, they seem to have some nice tools developed with it,

welcome once again,

VIJAY,

[Grumpus]

Most excellent stuff, Marin - thanks for the links. I'm going to play around in there some tomorrow!

G.

[Paul J]

I was wondering when Karon was going to chime in!

I'm still reading everything and will continue to do so. I can't imagine that keywords won't always be very important. Just upon first impression, a future and functional semantic web will need something to build on. I think keyword phrases in your tags and copy will always be relevant. "Copy that converts" suddenly makes a lot of sense now.

My brother in law said something recently that, in a way, makes sense to this topic. Not the exact words, but he said someting like - Google and other SE's will slow down their Keep the Faith When the [url=http://searchengineland.com/070621-145956.php]Keep the Faith When the Algo Changes[/url] when they understand what people are looking for when they search for "food".

Back to reading...

[copywriter]

Hi Paul,

I'm here Had some family stuff going on as of late and been amiss about posting. Please forgive me

Cool stuff, Marin and welcome I'll have to play around with those tools later on.

Another interesting aspect about semantic engines is that - unlike the current Google (or others) - the *more* information you include in your search, the more information you'll get in your results. Right now, the more information you include in a search the *less* results you receive.

But yes, keywords in copy will not be a thing of the past at all. They will still play a vital role in SEO copywriting.

According to my contacts, keywords will still provide the basis of the search and the search results. However... keywords within natural language will be most important. Simply cramming keywords into sentences where they make little sense won't do. You'll have to learn to incorporate keywords without destroying the flow of your copy.

This won't work very well:

Welcome to the best site for this, that and the other thing. We have the largest selection of this, that and the other thing of any site on the 'Net. Take some time and compare our prices on this, that and the other thing with other stores. You'll definitely come back and order this, that and the other thing from us when you see how low our prices are.

You'll have to make it flow better and sound more natural.

[Jill]

Gee...wonder where I've heard that before?

[Grumpus]

From me?

G.

[copywriter]

Yeah, no kidding. Which just goes to show that if you use long-term, common sense SEO techniques you'll have long-term high rankings.

[marin]

Hi,

Thank you very much for you kind welcome.
I am sorry to disappoint you, but I am not a scientist; I am just a site owner, who read a lot trying to survive after Florida. So I have no answers, I am just guessing.

After I discovered METALOG(very friendly software for "thinking" on the Web)
I understood that semantic web is totally different to the actual web, so Florida update has nothing to do with semantic or semantic web, but latent semantic seems to be appropriate.

Singular vs. plural: the ontology for singular is different from the plural one: try the “Near Neighbors” Colorado tool for restaurant (waiter waitress hamburgers chef waiters dinner) and then for restaurants (restaurants hotels stores establishments barbershops shops).
Suppose a small site “cheap restaurants in my city”; the site will cover the natural related terms and subjects from the singular ontology, but for sure will not include the hotels, shops etc. lists. Contrary, the big guys will build comprehensive site for the same city. Unique content vs. general content.

I used the One-To-Many Comparison Colorado tool to compare the sites in serp; first titles, then description and finally the body, but the results are contradictory.
For me the most intriguing is the allintitle ; there are big differences between gg ranking and One-To-Many Comparison Colorado evaluation.
How do you explain the actual allintitle?

[Jill]

Yeah, no kidding. Which just goes to show that if you use long-term, common sense SEO techniques you'll have long-term high rankings.

Ya think?

From me?

And you heard it from....?

:thumbsup:

[Grumpus]

And you heard it from....?

That'll be a chicken or the egg debate. I was fairly well into the game by the time I met you - but 'tis true, that you've been saying it as long as I've known you - which is - oh about a year and a half now, huh? Time flies!

G.

[OldWelshGuy]

An interesting year methinks.

We have semantic search, MS whaatever, yahoo, and best of all personalised search.

I am glad i am too dumb to understand all of this, so it means i don't have to worry about it