What is Latent Semantic Indexing?

• A method for finding latent patterns in unstructured data
• Based on a vector-space data model
• Core algorithm discovered 1874 (Beltrami, Jordan)
• First use in information retrieval in 1990 by Deerwester, Dumais, Furnas, Landauer, Harshman (Journal of the American Society for Information Science)
• Applications so far have been limited:

• Research in information retrieval
• Pedagody (automated essay grader)
• Mac OS 10.2 spam filter