Application of Inductive Logic Programming to Produce Emergent Behavior in an Artificial Society

• Published in: IIAI-AAI 2014 : proceedings : 2014 IIAI 3rd International Conference on Advanced Applied Informatics : Kitakyushu, Japan, 31 August - 4 September 2014 pp. 942 - 947
• Main Authors: Komura, Keigo, Mutoh, Atsuko, Inuzuka, Nobuhiro
• Format: Conference Proceeding Journal Article
• Language: English
• Published: IEEE 01.08.2014
• Subjects: Algorithms; artificial society; Classification; Computational modeling; Computers; Conferences; Educational institutions; Electronic mail; Emergence; emergent behavior; inductive logic programming; Learning; Logic programming; Machine learning; Representations; Search problems; Sugar
• ISBN: 1479941743; 9781479941742
• DOI: 10.1109/IIAI-AAI.2014.185

Artificial society is a discipline to study mechanisms of social system and phenomena which the mechanisms make. Emergence is global phenomena occurred by local mechanisms, such as, by collective behavior of autonomous agents. Understanding of emergence phenomena is a challenging subject. In this paper we use the framework of inductive logic programming (ILP) for artificial society and emergence behavior study. ILP is a branch of machine learning based on logic programming and inductive inference. We investigate the possibility of ILP in artificial society study. ILP and logic programming technique are applied to representation of an artificial society model, called Sugarscape, and to rule learning for agent behavior. Although classical ILP algorithms target classification problems, the proposed algorithm grows behavior rule for an evaluation measurement. Phenomena which this paper treate is limited but we showed that ILP technique can be applied to study in the field of artificial society.