Latent Semantic Indexing (LSI) is Telcordia patented information retrieval method. LSI can retrieve relevant documents even when they don't share any words with your query. LSI does this by an automatic statistical analysis of which terms occur in similar contexts. (See Deerwester, Dumais, Landauer, Furnas, and Harshman. "Indexing by latent semantic analysis." Journal of the American Society for Information Science, 1990, 41(6), 391-407.)
There are two major sections to the LSI searching interface:
Initially, only the query parameters section will be visible. After the first search is performed, both the query parameters section and the results section will be available.
LSI returns a rank-ordered list of best matching documents, terms, or both documents and terms. You can select documents you like from this list for relevance feedback searches or select terms from the list to supplement the query.
Contact lsi@research.telcordia.com for more information about LSI.