Artificial Business Intelligence: Scaling Beyond the Real World with Cyc and LarKC

The explosion of availability of free and open information resources following the emergence of the Web2.0 paradigm has widened the prospects for constructing real Artificial Intelligence solutions that are able to learn, to reason and to speculate. This talk discusses the general class of problems that should be solvable in the near term, in part by exploiting available knowledge, and in part by collaboration between people and machines.

In the last few years significant advancement has been achieved in semantic, knowledge and context technologies as well as in methods for knowledge management. These technologies are becoming especially effective when applied to the capture, formalization and automated reuse of knowledge. In particular, these techniques have been demonstrated by Cycorp in specific intelligence and medical domains. Equally, though they may be applied to problems of managing business complexity to provide ABI - Artificial Business Intelligence. The explosion of availability of free and open information resources following the emergence of the Web2.0 paradigm has widened the prospects for constructing real Artificial Intelligence solutions that are able to learn, to reason and to speculate.

In my talk I'll discuss the general class of problems that should be solvable in the near term, in part by exploiting available knowledge, and in part by collaboration between people and machines. I'll show some examples of partial solutions, and describe in some detail the components of a more complete solution. The discussion will focus on the issue of scaling AI techniques up to real applications, both in terms of very large, inferentially sophisticated knowledges bases, like Cyc, and in terms of techniques for web scale inference - the goal of the FP7 LarKC project.

Presenter(s): 
File(s)Size
active09_witbrock_abi_01[1].pptx14.17 MB