Autonomous Action Generation of Humanoid Robot from Natural Language
Communication between human and robot is necessary for an intelligent robot to be active in daily life. Conversation is the basis of communication. Most of the instructions given to a robot are related to some actions. This paper reports the method to generate the action of the humanoid robot by conversation. For this purpose, comprehension of action instruction written by natural language is necessary. The system to understand natural language for generating action is proposed. This system consists of a semantic understanding method to arrange input information, the knowledge base of vocabulary related to actions, the knowledge base of the parameter for robots to act, and association mechanism to handle a word, which is not known. The system is capable of understanding many input sentences by the association mechanism using Concept-Base and a degree of association.
KeywordsNatural Language Body Part Basic Action Humanoid Robot Semantic Meaning
Unable to display preview. Download preview PDF.
- 2.Kojima, K., Watabe, H., Kawaoka, T.: Concept-base Refining with Thesaurus and Logical Relations for a Word Association-system. In: Proc. of KES 2001, Part 2, pp. 1590–1594 (2001)Google Scholar
- 3.Watabe, H., Kawaoka, T.: The Degree of Association between Concepts using the Chain of Concepts. In: Proc. of SMC 2001, pp. 877–881 (2001)Google Scholar
- 5.Shinohara, Y., Watabe, H., Kawaoka, T.: A Conversation Semantic Understanding Method using the Commonsense Judgment System, IPSJ SIG Notes, NL-153, pp. 89–96 (2003)Google Scholar