Behavior Capture with Acting Graph: A Knowledgebase for a Game AI System

  • Maxim Mozgovoy
  • Iskander Umarov
Conference paper

DOI: 10.1007/978-3-642-25731-5_7

Volume 7108 of the book series Lecture Notes in Computer Science (LNCS)
Cite this paper as:
Mozgovoy M., Umarov I. (2011) Behavior Capture with Acting Graph: A Knowledgebase for a Game AI System. In: Kikuchi S., Madaan A., Sachdeva S., Bhalla S. (eds) Databases in Networked Information Systems. DNIS 2011. Lecture Notes in Computer Science, vol 7108. Springer, Berlin, Heidelberg

Abstract

Behavior capture is a popular experimental approach used to obtain human-like AI-controlled game characters through learning by observation and case-based reasoning. One of the challenges related to the development of behavior capture-based AI is the choice of appropriate data structure for agents’ memory. In this paper, we consider the advantages of acting graph as a memory model and discuss related techniques, successfully applied in several experimental projects, dedicated to the creation of human-like behavior.

Keywords

Behavior capture learning by observation case-based reasoning knowledge representation 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Copyright information

© Springer-Verlag Berlin Heidelberg 2011

Authors and Affiliations

  • Maxim Mozgovoy
    • 1
  • Iskander Umarov
    • 2
  1. 1.University of AizuFukushimaJapan
  2. 2.TruSoft Int’l Inc.St. PetersburgUSA