ITA Software's Technical Seminar Series

Talk Topic: Making Computing Meaningful and Meaning Computable
Ken Haase, MIT & beingmeta
December 9, 2009

Abstract

Meaningful computing involves processes of interpretation, inference, and imagination which reflect human meanings and purposes. Meaningful computing is the key to systems which can effectively partner with human beings and serve as "active media" connecting people across time, space, culture, and language. This talk will focus on the nuts and bolts of making meaningful computing scalable and efficient. In particular, I'll be talking about the latest internal incarnation of the FramerD platform (www.framerd.org) and its implementation and features at multiple levels. These include low-level optimizations for modern CPUs and effective SMP, database and inference mechanisms for seriously distributed applications, and programming lanuguage constructs for problems both profound (e.g. dealing with ambiguity) and prosaic (e.g. generating web content).

Bio

Ken Haase is a researcher and developer working on representationally rich computation. He earned his Ph.D. from the MIT Artificial Intelligence Laboratory in 1990 with a thesis on representational invention which was overseen by Marvin Minsky and Thomas Kuhn. During the 1990s, Ken was faculty at the MIT Media Laboratory and worked on media annotation and search, large-scale knowledge bases, natural language understanding, analogical representation, and interlingual computing. In addition, he created the initial version of an undergraduate program in Media Arts and Sciences and helped launch several spin-off laboratories around Europe. In the mid-00s, he served as acting director and senior scientist at Media Lab Europe in Dublin. He is currently at beingmeta, a company he founded in 2001 to commercialize his work at MIT on semantic media databases