Semantic Based Learning of Syntax in an Autonomous Robot

Abstract

-It is the goal of the Language Acquisition Group at the University of Illinois at Urbana-Champaign (LAR-UIUC) to build a robot that is able to learn language as well as humans through embodied sensorimotor interaction with the physical world. This paper proposes cognitive structures to enable an autonomous robot to learn the syntax of two-word sentences using its understanding of lexical semantics. Production rules of syntax in Chomsky Normal Form will be explicitly represented using a hidden Markov Model. Preliminary results, in the form of simulated experiments, show that these models can learn representations of syntax in this form.

An integrated three-stage model towards grammar acquisition

“…By exploiting such results, we can then test various algorithms of grammar induction dynamically on a robot under an interactive scenario. To our knowledge, while research in robotics on language learning that encompasses grammar is underway elsewhere [18]- [21], these studies focus on the learning architecture of an autonomous robot. Our research is unique in putting the robot in an interactive learning environment with human participants, acting as teachers, who are not imposed any predetermined restrictions for their speech.…”