Player behavior and personality modeling for interactive storytelling in games

Current video games use simple methods to deal with interactive narratives and the enormous variety of player types. In this paper, we propose a novel approach to interactive storytelling in games, in which the quests and the ongoing story are determined in view of individual personality traits and behavioral attitudes in a nondeterministic way. Our method starts the process employing a new technique to assess the player's personality traits according to the well-known Big Five model. These traits are then used by a nondeterministic planning algorithm to define adaptive goal hierarchies. In addition, an artificial neural network is trained to predict player behaviors in real-time, allowing partial-order planning operators to use player behaviors and personality traits as logical terms in their preconditions. With this approach, a richer individualized experience is provided to the player, while preserving consistency with the conventions of the chosen genre.