Played around a bit yesterday with it, and have to say, I was spectacularly underwhelmed. Admittedly, my geekiness is somewhat limited, so that may be ther reason.

My first impression? Nothing new here, folks... move along now.

Well you were having some probs getting yer head around it Doc. For starters get past aynonyms and certainly KW density. This type of semantic relevance is about 'concepts'. Which terms/words are common among documents over a given training set. And hey, anyone can make their own tool if so inclined;

http://www.google.com/search?q=LDAcodeANDC++,C,JavaMatlabcode.&sourceid=groowe&ie=utf-8&oe=utf-8

Will it tell you what Google is doing? Sadly no. Can it be used as a reference point? Possible.

I am actually kind of glad to see someone finally come out and (almost) say that keyword density is dead.

Well, keyword density died in like 1997. I thought that was obvious to most people.

Unfortunately not - there was at least one WordPress SEO product heavily marketed this year that relied on keyword density.

It also gave bonus points for outward linking every 250 words, though it wasn't particular about whether the links were affiliate links, all the same, or nofollowed.

Lots of people on Sphinn promoted it

Dead? Lol... gives the connotation it WAS ALIVE!! Anyway, Bill, most forms of modern semantic analysis is about establishing inferred concepts from a set of text. You set up some benchmarks (training documents) and establish what are the related terms in play or in the case of phrase based, even a ratio of related phrases on a 'good' document.

The larger risk here of course is not directly attributable to this development, but likely to be a problem none-the-less. And that is LDA Snake Oil. We are certain to start seeing a spate of SEO providers shilling 'Google LDA' as we have for the last 5+yrs with 'Google LSI' - a fire I have tried almost annually to put out.

Isn't this one of the major stumbling blocks to the standards conversation? That without more knowledge of what's going on inside the black box, we can't readily be certain about anything? Or at least at a semi-granular level?

Alas, the genie is out of the bottle and we shall have to watch and see now. The hills are alive... with the sound of LDA... (sing along if ya like)

Please, don't stone me, but doesn't using your keyword in the title help ranking for the term? And doesn't using the keyword in the title increase the 'keyword density' or whatever you call it in the title?

Sure there's no secret formula and no certian percentage of kwds on a page can be considered as the ultimate SEO standard. But it does help having some kwd density (at least a little above zero) doesn't it?

Well, keyword density died in like 1997. I thought that was obvious to most people.

Yes, it's never been a worthwhile measurement, imo.

I wonder why it's still included in the BruceClay toolset.