Joe Buzzanga, Product Manager, Elsevier Engineering & Technology
Don’t let the hip-hop product name fool you. The second example that Elsevier showed me yesterday to illustrate their efforts to build value-added services on top of the journal literature is not trivial, but it is definitely next-generational.
Taking advantage of semantic technololgies and natural language processing, illumin8, said product manager Joe Buzzanga, is targeted at the corporate R&D market. And, unlike it’s cousin 2collab (See post, below.), it does carry a price tag.
Joe described illumin8 as an "advanced semantic engine to enable intelligent discovery by trying to expose content in new ways."
Backed up by 3 million full texts, 36 million abstracts, 22 million patent records, and 7 billion web pages, illumin8 is, he said, not just a federated search engine that will unearth a lot of citations, but a method for discovering meaning via natural language processing.
Using a trade-secret algorithm (Joe calls it "the secret sauce"), Elsevier crawls billions of web sites and uses "thousands of rules," in search of "meaningful statements" for R&D purposes . . . statements relating to problems and solutions. When the system finds a query match, it extracts the pertinent facts and displays the answers in table format for the user to evaluate.
In what may be a contestant for understatement of the year, Joe described it as "not your normal search engine."
I couldn’t get Joe to tell me how many companies currently use the service, but he did say that uptake has been fast and pretty much concentrated in North America in Fortune 500 companies. Elsevier is now ramping up to roll the service out in Europe.
ITI VP, Content