Latent Semantic Indexing is a way for search systems to mathematically understanding and representing language based on the similarity of pages and keyword co-occurance. A relevant result may not even have the search term in it. It may be returned based solely on the fact that it contains many similar words to those appearing in relevant pages which contain the search words.

See also:

  • Quintura Search – free LSI type keyword research tool.
  • Patterns in Unstructured Data – free paper describing how LSI works
  • SEO Book articles on LSI: #1 & #2 (Google may not be using LSI, but they are certainly using technologies with similar functions and purpose.)
  • Johnon Go Words – article about how adding certain relevant words to a page can drastically improve its relevancy for other keywords