Saturday, March 24, 2007

Massive Uncertainty

Today I saw a phrase I had not seen before: "massive uncertainty." I saw this today in a call for papers for the "International Conference on Scalable Uncertainty Management(SUM-2007)." The expression "massive uncertainty" is very probably not new; it was just unfamiliar to me. But I like the phrase -- because it hypothesizes, not that some events are very uncertain, but that some problems of uncertainty are massive -- because, I assume, some problems of uncertainty involve large amounts of data, very extended argument, or both. I suspect that the problems of factual uncertainty that one finds in some or many trials are, in this sense, "massive."
  • Massive uncertainty, of course, can lead to great uncertainty -- or, perhaps more often (when extended argument is so extended that it is untractable), to diffuse uncertainty judgments. But possibilities such as these do not do away with the distinctness of massive uncertainty. For example, massive amounts of data and very extended argument can, in principle, sometimes generate the conclusion that the probability of a very granular event is very high or very low.)
  • The conference web site: http://www.umiacs.umd.edu/conferences/sum2007/

    The conference blurb states in part:

    The Scalable Uncertainty Modeling conference aims to bring [together] all those interested in the management of large volumes of uncertainty together, irrespective of whether they [presumably, large volumes of uncertainty rather than people] are in databases or in AI.

    No comments: