The Necessity of Separating Control and Logic When Grounding Language Using Neuroevolution (2009)
In this research we analyze the task of evolving a neural network to understand simple English commands. By understand we mean that the final agent will perform tasks and interact with objects in its world as instructed by the experimenter. The lexicon and grammar are kept small in this experiment. This type of work where semantics are based on an agent's perceptions and actions is referred to as language grounding. A substantial literature exists in this area, as the ability to have a robot or computer understand natural language instructions is one major goal of Artificial Intelligence, but very little has been done that grounds language in actions. We discover that grounding in actions requires a separation of logic and motor control.
Technical Report HR-09-05, Department of Computer Sciences, The University of Texas at Austin.

Yonatan Bisk Undergraduate Alumni ybisk [at] yo-tech com
OpenNERO OpenNERO is a general research and education platform for artificial intelligence. The platform is based on a simulatio... 2010