Natural language understanding for generating grasp actions

  • Authors:

  • Hirokazu Watabe;Seiji Tsuchiya;Yasutaka Masuda;Tsukasa Kawaoka

  • Affiliations:

  • Dept. of Knowledge Engineering & Computer Sciences, Graduate School of Engineering, Doshisha University, Kyoto, Japan;Dept. of Knowledge Engineering & Computer Sciences, Graduate School of Engineering, Doshisha University, Kyoto, Japan;Dept. of Knowledge Engineering & Computer Sciences, Graduate School of Engineering, Doshisha University, Kyoto, Japan;Dept. of Knowledge Engineering & Computer Sciences, Graduate School of Engineering, Doshisha University, Kyoto, Japan

  • Venue:
  • KES'07/WIRN'07 Proceedings of the 11th international conference, KES 2007 and XVII Italian workshop on neural networks conference on Knowledge-based intelligent information and engineering systems: Part I
  • Year:
  • 2007

Quantified Score

Hi-index 0.00

Visualization

Abstract

For achieving smooth communications between human and an intelligent robot, it is important that the robot understands the meaning of the sentences of natural language. Most of the instructions given to the robot are related to some actions. This paper reports the method to generate the action of the humanoid robot, especially grasp actions, by understanding natural language (Japanese) conversation. For this purpose, comprehension of action instruction written by natural language is necessary. The system that understands natural language and generates proper grasp action is constructed. This system consists of a semantic understanding method to arrange input sentences, the knowledge base of vocabulary related to objects to grasp, the knowledge base of the parameter for robots to act, and association mechanism to handle a word that is unknown. The system is capable of understanding many input sentences by the association mechanism using Concept-Base and a degree of association.