Context-based decision-making for virtual soccer players.

Romain Bénard (1) , Pierre de Loor (1, 2)

1 LISYC - Laboratoire d'Informatique des Systèmes Complexes
2 CERV - Centre Européen de Réalité Virtuelle

This article introduces a decision-making model for virtual agents evolving in dynamic and collaborative situations. Agents and humans have to collaborate in a virtual environment. In order to enhance the collaboration, the agent decision-making model is based on notions close to human ones. Those notions are context and case based reasoning. After an introduction of dynamic and collaborative situations, we present the notion of context and we give a definition adapted to our framework. The next part describes the decision making process. This one relies on the case identification thanks to a graph search algorithm. The last part of this document illustrates our purpose with an example taken from our application.

Domaines: Informatique; Intelligence artificielle [cs.AI]; Système multi-agents [cs.MA]; Synthèse d'image et réalité virtuelle [cs.GR]

Romain Bénard, Pierre de Loor. Context-based decision-making for virtual soccer players.. Context'07, 2007, France. pp.2-15. ⟨hal-00611049⟩