To put in a layman’s terms, latent semantic indexing (LSI) is a technique of indexing— analyzing, listing or categorizing certain keywords or phrases in the contents of different websites, books or documents in such a way that they have contextually and conceptually the same or related intent and meaning despite the various words utilized in them.
The technique used in latent semantic indexing aims at discovering the keywords in the text that carry a latent relationship in structure and usage. The idea behind the concept of LSI is to collect data that is conceptually akin in meaning and context to the search queries entered by the searchers in the search engines. The search outcomes may possibly, for that reason not share the specific words or phrases entered by the searcher.
For example, if you use the word ‘Saddam Hussein’, the search engine could return articles about the Gulf War, scenario in Kuwait or Iran, the elite force of the Iraqi despot, UN sanctions, oil fields in Iraq and a lot much more with out even mentioning the search word ‘Saddam Hussein’.