Juru, Watson, I Say, Juru!

February 1, 2011

Quite a heated discussion at lunch today. One of the goslings was raving about Watson. The Jeopardy demo convinced the engineer that IBM had the next big thing in search. A person can ask a question and right away get the answer. Wow. I thought that type of computer system only worked under carefully controlled conditions, in demos, or in motion pictures.

That’s why the goslings were agitated when I said, “It is TV. TV does almost anything—well, anything—for money.” I pointed out that the game shows 21 and the $64,000 Question took some liberties to boost ratings. Have TV times changed that much? I said, “I don’t think so.”

I supported my argument by mentioning Juru. Do you remember that gem from IBM. Here’s what my Overflight system spit out.

Juru is / was a full text search “library” that would make short work of “small and mid-sized corpuses.” Of course, “small” and “mid-sized” are rarely defined either by IBM or other search researchers. The idea was that Java made it easy to run Juru on any platform. Of course, today, I don’t think Juru would work in the Android or IOS environment, but some day maybe.

Juru asserted that the system would:

Support different document types
Make use of links just like our every tweakable PageRank-type systems
On the fly summaries of documents
Clustering
Nifty ways to keep the indexes small and, therefore, zippy.

You can get some info at this link. There is some additional color here:

I reminded the goslings that IBM rolls out search solutions as part of its global marketing efforts. More to the point, I asked the goslings which vendors’ search systems IBM resells. I did not hear the magic words Autonomy or Endeca. IBM once loved Fast ESP.

If you want search from IBM, what do you get today? A version of the open source search solution Lucene. Why? It works pretty well. Juru, Watson, Web Fountain, et al? Well, make up your own mind with some head to head testing. I won the argument and still had to pay for lunch. Honk.

Stephen E Arnold, February 1, 2011

Freebie

Written by Stephen E. Arnold · Filed Under Business strategy, Marketing, Natural language processing, News, Search, Semantic, Text analytics, Text processing

Comments

Comments are closed.

Search the site
Subscribe to Beyond Search
Feature archive
News archive

Stephen E. Arnold monitors search, content processing, text mining and related topics from his high-tech nerve center in rural Kentucky. He tries to winnow the goose feathers from the giblets. He works with colleagues worldwide to make this Web log useful to those who want to go "beyond search". Contact him at sa [at] arnoldit.com. His Web site with additional information about search is arnoldit.com.