digitalghost - 6:49 pm on Oct 23, 2008 (gmt 0)

>>Google bombing

Definitely not Google bombing. Google bombing is intended to make a site rank for semantically unrelated words or concepts. What I'm suggesting is using anchor text and semantically related words to make a site rank for related words and concepts.

Say for example you have a site about WWII but you don't want World War II as a phrase anywhere on the site. So you get some links from other sites with "Word War II" in anchor text.

Then it becomes all about the content and the concept of the site itself. So you have page titles like, "The German Invasion of Poland", The Bombing of Britain", The Attack on Pearl Harbor", The Battle of Midway", "The Truth About Iwo Jima", "The Russian Front", The Battle of Bastogne", etc. Then it becomes concept density and not keyphrase density.

A site as described above could rank for WW II and need not have the phrase mentioned on the site at all. Call it natural language ranking, semantic association, whatever. I call it "conversational optimization".

For example, in a conversation about Hitler (staying on the subject of WW II) between two people, the proper name, "Hitler" might not be mentioned at all, or rarely, yet the entire conversation could take place and anyone listening would know who the two people were talking about. If phrases like "beer hall putsch", "final solution", "Third Reich", "Germany's dictator", and "crystal nacht" pepper the conversation, discerning the subject of the conversation is easy even for our carbon-based processors. So what do you think computer algorithms can do with some simple relationship analysis?