Title: Building of commonsense knowledge for home robotic services: 'bring something' and 'tidy things up'

Authors: Weerachai Skulkittiyut; Haeyeon Lee; Ngo Lam Trung; Makoto Mizukawa

Addresses: Graduate School of Engineering, Shibaura Institute of Technology, 3-7-5 Toyosu, Koto-ku, Tokyo, Japan ' Partner Robot Division, Toyota Motor Corporation, 4-18 Koraku, 1-chome, Bunkyo-ku, Tokyo, Japan ' Department of Computer Engineering, School of Information and Communication Technology, Hanoi University of Science and Technology, 1 Dai Co Viet Road, Hanoi, Vietnam ' Department of Electrical Engineering, Shibaura Institute of Technology, 3-7-5 Toyosu, Koto-ku, Tokyo, Japan

Abstract: Sharing commonsense knowledge would be one of the most important keys for interaction between human and service robot. However, the complication of human commonsense makes extracting appropriate information for service robot difficult. This paper presents our approach to build the basic-level knowledge network (BKN) as a commonsense and how it is used for robotic services. In this paper, we focused on two services. One is 'bring something' and the other is 'tidy things up'. We found that for 'bring something', relationship between object-activity is important, and for 'tidy things up' service, knowledge about relationship between object-place is more important. To perform these services at the practical level, many relative knowledge should be prepared. Moreover, we would like to introduce how BKN is built and show the demonstration result in which how BKN is useful for actual services in real environment.

Keywords: commonsense knowledge; basic-level knowledge networks; BKN; bringing; fetching; robotic services; tidying up; home robots; knowledge sharing; human-robot interaction; HRI; service robots; relative knowledge.

DOI: 10.1504/IJMA.2014.059770

International Journal of Mechatronics and Automation, 2014 Vol.4 No.1, pp.1 - 16

Available online: 07 Mar 2014 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article