Latent semantic indexing is a detailed analysis of the meaning of every word and term in an internet document. Latent semantic indexing is performed by a search engine in order to help sort articles by their relevance to a search query that a user enters into the search engine. Most words have more than one meaning. That is part of what makes language so complex. When a user does search query, they want to find the articles that are most relevant to their search, so a search engine needs to come up with a score for how well each article relates to the search term.
Latent semantic indexing is an import part of computing the overall score of an article. The word "semantic" is the technical term for the meaning of words and "indexing" means that the semantic value of the words or terms will be arranged in a ranked list. The word "latent" is a vital part of this term because words do have more than one meaning, and the meaning of a word can change depending on the other words surrounding it. For example, the word "bear" probably refers to a large furry mammal you would find in the woods or a zoo. But a "bear market" has little to do with the mammal, other than the fact that most people run from it.
That idea of running from both types of bear is an example of the latent meaning of "bear market". That's why latent semantic indexing is important. Latent semantic indexing performs a detailed semantic analysis on all the words and terms in every article on the internet, and ranks the different possible meanings for key terms. Finding the word "market" after "bear" decreases any value assigned to the "mammal" definition of bear, but maintains the value of the "run away" portion. This is also called value decomposition, which is part of the nitty-gritty of how latent semantic indexing works.
For the purposes of internet marketing, of course, you don't need to know all the details of how latent semantic indexing works, but this background should help you be sensitive to what it is doing. What you will need to know is how to get your articles ranked highest on key terms that users search for. You can see that this will require some analysis on key search terms and related words. There are web sites that provide this information. Then, load your articles with the key words and related terms, and they will rise to the top of every query.