At Inxight, we have a Google add-on product that, among other things, can provide insight into the full-text content of documents at the summary stage.
It shows you the most relevant people, companies, places, etc., that are mentioned in the full text of the document.
This provides a better preview into the document's true contents. It lets me do several tasks more effectively, like competitive research and expert location.
For example, a typical Google keyword-in-context summary (Google desktop search for "linguist" looks like this:
Linguistics Symposium in 14th year.doc
said Davis. We are not just grammarians. While you certainly have to have an appreciation for many languages, being a linguist doesn’t mean knowing many
An Inxight-enhanced summary looks like this:
Linguistics Symposium in 14th year.doc
said Davis. We are not just grammarians. While you certainly have to have an appreciation for many languages, being a linguist doesn’t mean knowing many
Top Mentions: Alan Kaye; Colleen Davis; Elizabeth Closs Traugott; Stanford University; Stefanie Calvillo
How do we do this? Inxight has software that "reads" text in 32 different languages and, based on linguistics and patterns, can detect 35 different "entities" in that text -- people, companies, organizations, dates, currencies, concepts (like "unstructured data" or "global piracy"), etc.
I think it's pretty neat, and although it's not perfect, I'm still impressed by it daily.
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment