Michael Mateas
Location: (Santa Cruz, California)
Personal Research Web Page: http://www.soe.ucsc.edu/~michaelm
Keywords: Interactive Storytelling, Game AI, Procedural Content Generation, Authoring Tools, Drama Management, Story Generation, AI-based Art
Posted on: Tuesday, May 18th, 2010
Broad Research Area: AI / Machine Learning / Robotics / Vision, HCI / CSCW
Research Interests:
The Expressive Intelligence Studio (EIS) at UC Santa Cruz has four affiliated faculty: Michael Mateas, Noah Wardrip-Fruin, Arnav Jhala and Jim Whitehead. We do fundamental technical and design research to enable new kinds of interactive art and entertainment forms. Current research directions in EIS include automatic content generation, interactive storytelling, and advanced game authoring tools.
Automatic content generation. The current state of the art for game development involves the hand-creation of much of the content in a game, such as hard-coded scripts, level designs, models, sounds, and animations. As hardware capabilities and consumer expectations increase, ever larger team-sizes are needed to produce the number and quality of assets required, resulting in ever more expensive production costs. Automatic content generation resolves this production crisis in the games industry; additionally it enables games that dynamically adapt to the specific play styles and goals of individual players as well as making it possible for high-quality games to be produced on much smaller budgets, broadening game creation. Current EIS research in content generation includes generation of quests, based on an analysis of gameplay history, level generation for platform games, and scenario generation driven by learning goals for training games.
Interactive Storytelling. Games have the potential to be an incredibly powerful storytelling medium, combining the cathartic pleasures (and reflection on the human condition) of well-formed stories with the exploratory power and high-agency experience of interaction. Game-based storytelling, however, is still in its infancy. In almost all games and virtual worlds, the story itself does not change as a function of the choices the player makes; there are no game mechanics at the story-level that make the story itself playable. Fundamental research is required to enable the creation of rich, autonomous characters; dynamic, compelling storylines; and intelligent, automatic cinematography—so that character interactions and story arc depend on player actions, yet are presented as compellingly as today’s scripted experiences. EIS research in interactive storytelling builds on the foundations of the award-winning interactive drama Façade (www.interactivestory.net), to create next-generation autonomous characters and drama-management technologies.
Advanced Game Authoring Tools. Game designers currently lack software-based tools to facilitate the design process. While game engineers have game engines and middleware libraries to bootstrap game implementation, and artists have tools to facilitate the creation of 3D models, textures, and animations, designers have no tools that explicitly support their specialized work. In the early stages of professional game design, designers work out the interactions and implications of their game mechanics by building stripped-down playable prototypes, either computationally or with physical materials. But gaining design insights from such prototypes, particularly insights regarding game rule interactions, is notoriously difficult, even for expert designers. Further, there is increasing interest in supporting game development novices in creating game content, including supporting domain experts in creating serious games. Current EIS research in this area includes intelligent game prototyping tools that can reason about first-class models of game mechanics, design brainstorming tools for quest design, and storytelling tools that help novice authors create interactive stories.
Contact Information:
Please email me at email obfuscated - click to reveal.edu
