LSI vs Vector Keywords

LSI Keywords


Discover exactly what kind of keywords you should be using for SEO.


There has been a lot of talk in the SEO industry of late about LSI keywords. They are supposed to increase your websites relevance by including other idustry based terms. And they do, to some degree. But LSI is by far the least effective way to increase relevance and rankings. LSI was developed to assign an overall category to a page that displays Adsence ads. Not to drill down on specific search queries.


However 3D vector keywords (which are essentially reverse LDA) cut to the core of how search engines and information retrieval systems assign reverence and rankings.


This is an extract from the conclusion of a recent post on document relevancy:


‘This type of query-document similarity scoring is well established in the research literature and underlies every modern information retrieval system. As such, it is fundamental to search and is immune to algorithm change.’


The basic principal is every word in a document is assigned a probability that it’s related to a certain topic or search vector (query). So if you have an article on dog training, but your don’t mention any other relevant terms and start talking about how how great your product is you are going to score a very low relevance, and not rank anywhere near as good as if you had a highly relevant document. Don’t forget your can still sell your product fine, you can minimize text using scrolling and expanding boxes. These must be used in a white hat way, displayed clearly. Hidden text will not help you.


Here are examples of LSI keywords, which are found at the bottom of any search result.




As you can see these are just broad category type keywords. If you were a search engine and you were trying to assign the probability of a word being relevant to an over all document on dog training, do you think that would be a good way to do it? Vector keywords on the other hand posses far more power, here are some example of vector keywords.





As you can see these words are not simple relevant sub topics, they make up the core of how the engines asses the probability of each word relating to the search vector query. They are extremely powerful and are used in SEO to boost rankings through the roof, cast a far wider long tail net and reduce the amount of content you actually need to write. Relevant documents rank for 100’s sometimes 1000’s of keyword searches because the effects of relevance compound on top of each other.


Relevance breeds relevance for other close searches. Where as ‘targeted’ content that targets just one search phrase using keyword density will likely only rank for 1 search phrase, and you’ll need a lot more links to get it there. It will also stand you in the firing line of penlites as it’s completely unnatural to have 10 documents all targeting slightly different keywords.


Unlock the power of 3D vector relevance now.


Sign Up Now

  • What about the combination of LSI and Vector Keywords and how they influence search rank as opposed to one versus the other?