Generalised Player Modelling : Why Artificial Intelligence in Games Should Incorporate Meaning, with a Formalism for so Doing

Show full item record



Permalink

http://hdl.handle.net/10138/319281

Citation

Cowley , B U 2020 , Generalised Player Modelling : Why Artificial Intelligence in Games Should Incorporate Meaning, with a Formalism for so Doing . in X Fang (ed.) , HCI in Games : Second International Conference, HCI-Games 2020, Held as Part of the 22nd HCI International Conference, HCII 2020, Copenhagen, Denmark, July 19-24, 2020, Proceedings . Lecture Notes in Computer Science , vol. 12211 , Springer , Cham , pp. 3-22 , International Conference, C &C 2020, Held as Part of the 22nd HCI International Conference, HCII 2020, Copenhagen, Denmark, July 19–24, 2020 , Copenhagen , Denmark , 19/07/2020 . https://doi.org/10.1007/978-3-030-50164-8_1

Title: Generalised Player Modelling : Why Artificial Intelligence in Games Should Incorporate Meaning, with a Formalism for so Doing
Author: Cowley, Benjamin Ultan
Other contributor: University of Helsinki, Department of Education
Fang, Xiaowen



Publisher: Springer
Date: 2020-07-10
Language: eng
Number of pages: 20
Belongs to series: HCI in Games Second International Conference, HCI-Games 2020, Held as Part of the 22nd HCI International Conference, HCII 2020, Copenhagen, Denmark, July 19-24, 2020, Proceedings
Belongs to series: Lecture Notes in Computer Science
ISBN: 978-3-030-50163-1
978-3-030-50164-8
DOI: https://doi.org/10.1007/978-3-030-50164-8_1
URI: http://hdl.handle.net/10138/319281
Abstract: General game-playing artificial intelligence (AI) has recently seen important advances due to the various techniques known as ‘deep learning’. However, in terms of human-computer interaction, the advances conceal a major limitation: these algorithms do not incorporate any sense of what human players find meaningful in games. I argue that adaptive game AI will be enhanced by a generalised player model, because games are inherently human artefacts which require some encoding of the human perspective in order to respond naturally to individual players. The player model provides constraints on the adaptive AI, which allow it to encode aspects of what human players find meaningful. I propose that a general player model requires parameters for the subjective experience of play, including: player psychology, game structure, and actions of play. I argue that such a player model would enhance efficiency of per-game solutions, and also support study of game-playing by allowing (within-player) comparison between games, or (within-game) comparison between players (human and AI). Here we detail requirements for functional adaptive AI, arguing from first-principles drawn from games research literature, and propose a formal specification for a generalised player model based on our ‘Behavlets’ method for psychologically-derived player modelling.
Subject: 515 Psychology
6162 Cognitive science
113 Computer and information sciences
Rights:


Files in this item

Total number of downloads: Loading...

Files Size Format View
Cowley_HCIG2020.pdf 434.5Kb PDF View/Open

This item appears in the following Collection(s)

Show full item record