eQuery is a Natural Language Processing-based automatic information
access system which implements algorithms that interpret language
at all the levels at which humans are known to extract meaning.
These are the morphological, lexical, syntactic, semantic, discourse
and pragmatic levels. Using these levels eQuery extracts important
concepts and relations from texts (both documents and queries)
by applying sophisticated and proven natural language processing
techniques.
The information retrieval capabilities exist in a 2-stage
retrieval system, in which the Language-to-Logic (L-2-L) module
is used to process the user's initial query into the richest,
fullest query representation possible. This representation
is then fed into a query converter for each search engine
which, in turn, produces the richest representation usable
by the selected COTS Web Search Engine(s). The search engine
does its full web search and the top N-selected documents
are returned and run through eQuery's full NLP document analyzer
module. The full L-2-L query representation is now utilized
and matched against the representation of these top-ranked
documents that have been processed through the full NLP system,
and a re-ranked list of documents is presented to the user
or alternatively, to a selected visualization tool.
|
