Latent semantic analysis (LSA) is a theory and technique that examines the relationships between sets of documents and the terms they use. Latent semantic indexing (LSI) uses mathematical methods to determine the relationship between words people use in search queries and concepts in web content. LSI leverages the theories of LSA in order to increase SEO.

It works like this: a webpage’s contents are crawled by Google, Bing, Yahoo! or other search engines for common words or phrases. Search engines collect those words and phrases and identify them as keywords for the page. Though often used interchangeable, LSA is the theory of analysis, whereas LSI is the application of that theory.

LSI can determine the similarities in the meanings of words to one another. For instance, if a page is titled, “Plumbing Help,” search engines would expect words like toilet, sinks, and pipes to show up as well. LSI assumes that words and phrases close in meaning will appear in similar types of texts. LSI is useful for finding relationships between terms, web page grouping, and search engine optimization (SEO) through using the right keywords and associated keywords.