A document indexing process, used by Search Engines, that records which keywords a document contains and examines the document collection as a whole to see if any other documents contain the same keywords. LSI considers documents that have many common words to be semantically close, and those with fewer in common to be semantically distant. When an LSI-indexed database is searched it looks for similitarity values it has calculated for every content word, and returns documents it thinks best fit the query. LSI does not require an "exact match" to return useful results because two documents may be semantically close even if they don't share a particulary keyword.