Learning from Play: Facilitating character design through genetic programming and human mimicry

Swen Gaudl, Joseph Osborn, Joanna Bryson

Research output: Contribution to conferencePaper

  • 1 Citations

Abstract

Mimicry and play are fundamental learning processes by which
individuals can acquire behaviours, skills and norms. In this paper we
utilise these two processes to create new game characters by
mimicking and learning from actual human players. We present our
approach towards aiding the design process of game characters through
the use of genetic programming. The current state of the art in game
character design relies heavily on human designers to manually create
and edit scripts and rules for game characters. Computational
creativity approaches this issue with fully autonomous character
generators, replacing most of the design process using black box
solutions such as neural networks. Our GP approach to this problem not only mimics actual human play but creates character controllers which can be further authored and developed by a designer.
This keeps the designer in the loop while reducing repetitive labour. Our system also provides insights into how players express themselves
in games and into deriving appropriate models for representing those insights. We present our framework and preliminary results supporting
our claim.

Conference

Conference17th Portuguese Conference on Artificial Intelligence (EPIA-2015)
CountryPortugal
CityCoimbra
Period8/09/1511/09/15

Fingerprint

Genetic programming
Personnel
Neural networks
Controllers

Keywords

  • agent design
  • games
  • genetic programming
  • machine learning

Cite this

Gaudl, S., Osborn, J., & Bryson, J. (2015). Learning from Play: Facilitating character design through genetic programming and human mimicry. Paper presented at 17th Portuguese Conference on Artificial Intelligence (EPIA-2015), Coimbra, Portugal.

Learning from Play : Facilitating character design through genetic programming and human mimicry. / Gaudl, Swen; Osborn, Joseph; Bryson, Joanna.

2015. Paper presented at 17th Portuguese Conference on Artificial Intelligence (EPIA-2015), Coimbra, Portugal.

Research output: Contribution to conferencePaper

Gaudl, S, Osborn, J & Bryson, J 2015, 'Learning from Play: Facilitating character design through genetic programming and human mimicry' Paper presented at 17th Portuguese Conference on Artificial Intelligence (EPIA-2015), Coimbra, Portugal, 8/09/15 - 11/09/15, .
Gaudl S, Osborn J, Bryson J. Learning from Play: Facilitating character design through genetic programming and human mimicry. 2015. Paper presented at 17th Portuguese Conference on Artificial Intelligence (EPIA-2015), Coimbra, Portugal.
Gaudl, Swen ; Osborn, Joseph ; Bryson, Joanna. / Learning from Play : Facilitating character design through genetic programming and human mimicry. Paper presented at 17th Portuguese Conference on Artificial Intelligence (EPIA-2015), Coimbra, Portugal.6 p.
@conference{2ffbb5ecbfba428faa05d6ad655c021e,
title = "Learning from Play: Facilitating character design through genetic programming and human mimicry",
abstract = "Mimicry and play are fundamental learning processes by whichindividuals can acquire behaviours, skills and norms. In this paper weutilise these two processes to create new game characters bymimicking and learning from actual human players. We present ourapproach towards aiding the design process of game characters throughthe use of genetic programming. The current state of the art in gamecharacter design relies heavily on human designers to manually createand edit scripts and rules for game characters. Computationalcreativity approaches this issue with fully autonomous charactergenerators, replacing most of the design process using black boxsolutions such as neural networks. Our GP approach to this problem not only mimics actual human play but creates character controllers which can be further authored and developed by a designer.This keeps the designer in the loop while reducing repetitive labour. Our system also provides insights into how players express themselvesin games and into deriving appropriate models for representing those insights. We present our framework and preliminary results supportingour claim.",
keywords = "agent design, games, genetic programming , machine learning",
author = "Swen Gaudl and Joseph Osborn and Joanna Bryson",
year = "2015",
month = "9",
day = "8",
language = "English",
note = "17th Portuguese Conference on Artificial Intelligence (EPIA-2015) ; Conference date: 08-09-2015 Through 11-09-2015",

}

TY - CONF

T1 - Learning from Play

T2 - Facilitating character design through genetic programming and human mimicry

AU - Gaudl,Swen

AU - Osborn,Joseph

AU - Bryson,Joanna

PY - 2015/9/8

Y1 - 2015/9/8

N2 - Mimicry and play are fundamental learning processes by whichindividuals can acquire behaviours, skills and norms. In this paper weutilise these two processes to create new game characters bymimicking and learning from actual human players. We present ourapproach towards aiding the design process of game characters throughthe use of genetic programming. The current state of the art in gamecharacter design relies heavily on human designers to manually createand edit scripts and rules for game characters. Computationalcreativity approaches this issue with fully autonomous charactergenerators, replacing most of the design process using black boxsolutions such as neural networks. Our GP approach to this problem not only mimics actual human play but creates character controllers which can be further authored and developed by a designer.This keeps the designer in the loop while reducing repetitive labour. Our system also provides insights into how players express themselvesin games and into deriving appropriate models for representing those insights. We present our framework and preliminary results supportingour claim.

AB - Mimicry and play are fundamental learning processes by whichindividuals can acquire behaviours, skills and norms. In this paper weutilise these two processes to create new game characters bymimicking and learning from actual human players. We present ourapproach towards aiding the design process of game characters throughthe use of genetic programming. The current state of the art in gamecharacter design relies heavily on human designers to manually createand edit scripts and rules for game characters. Computationalcreativity approaches this issue with fully autonomous charactergenerators, replacing most of the design process using black boxsolutions such as neural networks. Our GP approach to this problem not only mimics actual human play but creates character controllers which can be further authored and developed by a designer.This keeps the designer in the loop while reducing repetitive labour. Our system also provides insights into how players express themselvesin games and into deriving appropriate models for representing those insights. We present our framework and preliminary results supportingour claim.

KW - agent design

KW - games

KW - genetic programming

KW - machine learning

M3 - Paper

ER -