Standardizing the Representation of User Tasks

Currently, most demonstrationai systems are built from scratch, typically by individuals, and are all quite unique and distinct from each other. It is still desirable to build more such small-scale systems in the future. However, the key to a larger-scale, collaborative and interdisciplinary effort is in defining a common representation for user tasks, and in defining a common input format for automation tools. This way, the Machine Learning community can focus on the inferencing while the Programming By Demonstration community can focus on the human interface for automating tasks, using the same representation. This paper consists of two parts. The first section makes a case for spending some time at the symposium designing a common representation for user tasks and their automation. The second section briefly presents Grizzly Bear, a demonstrational tool that I have built over the last three years (and which taught me the importance of a well-defined underlying representation the hard way). 1 Position Statement or "What we should be spending our time on at the Symposium" I believe that the key to a wider collaboration on automating tasks for users is in agreeing on a common format for internally representing these tasks. Therefore, I think it is worth spending some time at the symposium on this issue, and to maybe form a "standards sub-committee" that could keep working afterwards.