KW Density Not Used as FarBack as 1997

<removed>

Hello there,

In this regard, I would comment here keyword density has been changed with the another name. Currently, keyowrd density matter but only according to page content value. How and what quality you are giving to your webpages in site. Interlinking is new procedure for SEO area. This techniques is being used to increase quality of your keyword density which is read fastly by search engines. Interlinking with internal pages in right keywords.

The previous keyword density procedure has gone and there is no consideration in search engine for that. Andi it also depends upon the search engines techniques because there many search engine and all have their own techniques.

Surender Singh Negi

So, it seems that they might flatten a collection of terms such as photo, photograph, picture down to a single entity. And having ANY of these terms on your page gives you some automatic points towards all of them. I have seen this for my own sites where I have not a single use of the term on my page and not a single use of the term on any backlinks, yet the page shows in the SERPs for that term. IMO you can learn the most about this - as it applies to your own sites - by studying your logs and see where you are getting traffic from related terms that do not appear on your page.

I do think that it will still pay you to do keyword research and use the term (photo, photograph, picture) that best applies to your page or that might bring the greatest traffic - as IMO this effect is weaker for terms that are not on your page.

Negi - I'm afraid I cannot understand your point. Perhaps you could clarify?

EGOL - What's great about stemming and tokenization is that it takes words in your copy and reduces them to a singular format to determine subject matter and conduct topic identification.

So, in the example above, "flew", "flies" and "flown" all become fly through stemming, but in tokenization, "computer", "computational" and "computing" all become "comput" - which is then mapped to a topic area.

If we can ever get a tool going, this will be awesome.

Feeling kinda exhausted here, so just a quick note - As far as the "comput" goes, it does depend on what stemmer you decide to use. You can have a stemmer which works on heuristics depending on pos-tags.

Topic area has been defined by wordnet in the past - it works to a point, it is not an exact science, but then what is! You would get a lot of converging ones so you might then want to form semantic groups. Also one word may not mean the same thing in another context.

Exact - KW density and weighting and so on has been going on for a long time. It has gone on in fact much longer than since 1997.

Things have got a lot more complex than that even. NLP is huge and still growing, and seems to be converging with many other areas of study. The things going on now go beyond the words on the page, but map everything to a different dimension if you like, because things like author intention and reader intention are used.

Just a quote I like:

"Computer science is no more about computers than astronomy is about telescopes."
(Edsger Dijkstra)

Fascinating. Great find as usual Randfish!

I would love a tool that would suggest keywords if I gave it a specific keyword(s) based on this technology Would really help me optimize my pages - hehehe...

G-Man

Xan - Thanks for stopping by and verifying all this. I'm always galad to have someone to check my facts!

Geoff - I certainly want to try to build something like this... We'll see what can happen, though.

That's some really nice findings, Randfish! But, developing a tool like that, wont be as easy as it seems... ;)

Nah, you'll be fine. Keep your head down Rand!