Researchers frequently lament the predictable framework of published papers.  The constraints of a rigorous review process discriminate against unconventional work and ideas that are innovative but not yet fully worked out.  As part of its mission to identify major new research opportunities, the CCC is sponsoring a series of “wacky ideas” sessions at several conferences. The goal of these sessions is to break free of the shackles of the normal reviewing process while still requiring a paper.  In this way, the “wacky idea” sessions differ from a “midnight session” of informal talks, in that the paper allows the ideas presented to be more broadly accessible.

The first of these sessions were held last summer at PLDI and last fall at OSDI.  Now, the CCC has sponsored a third session, this time at the Conference on Innovative Data Systems Research (CIDR).  The CIDR conference, which began in 2003, is a database conference that seeks papers about innovative and risky approaches, systems-building experience, killer applications and “war stories,” experimental studies, unsolved technical challenges, provocative position statements, and other off-the-beaten-path papers on the architecture and implementation of data-centric systems.  As such, the conference itself already goes part way towards addressing the CCC’s “wacky ideas” initiative.  Nevertheless, for the 2011 conference, CIDR added a CCC-sponsored “Outrageous Ideas and Visions” (OIV) track that focused on long-term challenges and opportunities for the database and data management communities that are outside of current mainstream research in the field.  Of particular note, the OIV track allowed for papers that lack any grounding in current or near-term practice or that address issues currently getting insufficient attention from the field. » Read more: “Outrageous Ideas” at CIDR: Seeking to Stimulate Innovative Research Directions

“What is Jericho?”

Those were the first words from “Watson,” the IBM supercomputer system that’s taking on Ken Jennings and Brad Rutter — the two winningest players in “Jeopardy!” history — this week.

Minutes later, with three categories of questions completed as part of this morning’s dry run, Watson was winning:  the supercomputer had $4,400; Jennings trailed with $3,400; and Rutter was third with $1,200.

If those first few minutes are any indication of what the actual game shows (which will be taped beginning tomorrow) are going to be like, we could be in for a truly fascinating man v. man v. machine matchup when the shows hit the airwaves February 14-16.

As we’ve covered in this space before (here and here), “Watson” is the result of years of research in areas like natural language processing, AI/machine learning, and so on.  As described in an Engadget.com article recounting this morning’s sneak preview:

Watson has thousands of algorithms it runs on the questions it gets, both for comprehension and for answer formulation. The thing is, instead of running these sequentially and passing along results, Watson runs them all simultaneously and compares all the myriad results at the end, matching up a potential meaning for the question with a potential answer to the question. The algorithms are backed up by vast databases, though there’s no active connection to the internet — that seems like it would be cheating, in Jeopardy terms.

Much of the brute force of the IBM approach (and why it requires a supercomputer to run) is comparing the natural language of the questions against vast stores of literature and other info it has in its database to get a better idea of context — it has a dictionary, but dictionary definitions of words don’t go very far in Jeopardy or in regular human conversation. Watson learns over time which algorithms to trust in which situation (is this a geography question or a cute pun?), and presents its answers with a confidence level attached — if the confidence in an answer is high enough, it buzzes in and wins Trebek Dollars.

Be sure to check out video from this morning’s event!

(Contributed by Erwin Gianchandani, CCC Director)