Hi can anyone help me uderstand LSI Latent Symentic Indexing? I read some articles but it looks like too much complex for beginner like me. Please brief me how it works and how it helps in SEO. how do i benefit from it.
I'm not exactly sure what LSI is either I never took the time to fully understand it, but some of it has to do with relavent or related search terms. For instance, if you search for the keyword "beds" using the tld key "~beds" you'll see related terms such as , bedding, linen, etc. that's bolded as well. I try to incorporate them into the page along with the main keyword because Google sees them as related making the page more relevant. I'm using a tool (keyword analyzer, I think that's the name) that searches the first page of Google and lists the commonly used terms on those pages and it's working surprisingly well, it's not a "push button" kind of thing though.
It's a mathematical way of evaluating the relationship of all words to all other words. Anyone who tells you they understand it is probably fooling themselves, because it looks like this - http://nlp.stanford.edu/IR-book/html/htmledition/low-rank-approximations-1.html generally when people refer to lsi, they actually mean a much cruder semantic/synonymatic way of evaluating the meaning of words in relation to other words, such as google use. (best observed in their use of broad match in adwords - which also shows how inaccurate it can be)
It's rumored that Google and other engines might use this to evaluate relavance. For instance, if someone searches at Google for "car maintenance", Google is not only considering pages that have <title>, <h1>, <h2>s, content, etc. containing "car maintenance", "car", or "maintenance"... They are not only looking at pages with inbound links containing "car maintenance", "car", or "maintenance"... Or pages that have links to them from other pages with "car maintenance", "car", or "maintenance" on the referring page. They are ALSO considering pages, links, and pages with refering pages that contain words like "vehicle", "auto", "motor", "BMW", "automotive", "car care", "care", "cleaning" because they can determine through Lexical Semantic Analysis that these words are all very closely related to one another. When "car" or "maintenance" appears in content, words like "vehicle", "auto", "motor", "BMW", "automotive", "car care", "care", "cleaning", etc. also seem to appear frequently in close proximity to "car" and/or "maintenance". So they figure they are related to one another... This is a WAAAAAAAAAAAY oversimplified explanation, but hopefully it helps picture what is going on. Of course, having exact matches on "car maintenance" carry the most weight toward helping your rank for that phrase... but having "car", "maintenance", "vehicle", "auto", "motor", "BMW", "automotive", "car care", "care", "cleaning", etc on the page being considered, in it's inbound link texts, on referring pages will ALL help you rank for the term "car maintenance". Using these Lexically Semantic "related" words in your content gives you more words to use in your content to help you rank for a phrase (though it helps to a lesser degree) without it sounding spammy. Since all the words are related and help each other rank, it helps your pages also rank for phrases and words that you might not even have targeted. With all the PhDs at Google, I'm sure "they" understand it totally. They probably have done a lot to advance the science of LSA. There is an operator you can use at Google to see what words they consider related by LSA/LSI. It's the ~ (tilda) operator. So if you search for "~car maintenance" (without the quotes) you'll see that they highlight words like car, vehicle, auto, BMW, motor because they are all words closely related to "car". If you search for "~car ~maintenance" (without the quotes) you'll start to see words like "care", "cleaning", etc. which are related to maintenance begin to be highlighted along with the "car" related terms. I always keep my Google preferences to 100 results per page... so it's easy to scroll down and see lots of variations.
only this much I knw...LOL thanx for this...wld surely help to minutely optimize the most important pages
It's all the descriptive words around the keywords that build the importance of the keyword - isn't it? Eg soft pillow, fluffy pillow (pillow being the keyword) light feather duvet that allows your body to breathe, silky soft Egyptian cotton duvet cover,