Research

 

Current Active Studies


    DASSIEs Study 1 - Online Study being conducted until June 2009


Current Areas of Research

    Interactive AI (Computational Intelligence)

  1. •Automatic Spatial Decomposition Methods for Agent Navigation

  2. •Symbolic Annotation and Reasoning on Simulation Objects

  3. •Geometric Spatial Knowledge Development in Interactive Worlds

  4. •Visual Data Mining of Player Traces in Interactive Environments

  5. •Understanding Learning and Behaviors in Interactive Environments

  6. •User-generated Behavior-based Control Agents for Simulation Training

  7. •Using Inferential Geometry to Localize Simulation Events in Network Streams

    Simulation Learning

  1. •Games and Simulations to Promote Learning

  2. •Advancing Computer Science Education Using Games

       

Current Major Project Initiatives

   

    DASSIEs (Dynamic Adaptive Super-Scalable Intelligent Entities)

       User-created agents using a behavior-based control architecture for rapid deployment in mission rehearsal and other ad hoc simulation training. This work explores the intelligent user interface for common warfighter creation of agents as well as the core percepts, actions, and infrastructure for realizing those agents in 3D polyhedron-based virtual worlds. This work is supported by a grant from DARPA.


    CGUL (The Common Games Understanding and Learning Toolkit)

        The CGUL (Common Games Understanding and Learning, pronounced “seagull”) Toolkit that is broken down into three parts: Data Providers and Assistance Tools, Information Support Services, and Evaluation Services. These tools have been designed to work with any FPS or third-person shooter (3PS) game using polyhedron-based environments. The Data Providers and Assistance Tools provide a number of information creation elements to improve an environment for player experience, game AI usage, and interaction analysis. Advanced spatial decomposition, symbolic annotation of the environmental elements, calculating the information value of the surfaces in an interactive environment, and geometric standardization and analysis form the core. The Information Support Services provide in-game information to agents and the game while capturing data for dynamic and post-game analysis. The Evaluation Services focus on how logging player-centric game data can be used to better understand both human and artificial player behavior and learning through the use of visual data-mining, graph-based interaction representations, collected metric analysis, and clustering tools.


Research Affiliations


    Director, Game Intelligence Group at UNC Charlotte

    Co-Director, Games + Learning Lab at UNC Charlotte

Research Interests


In general, my research interests are in discovering complimentary artificial intelligence (AI) techniques that are more powerful in combination than by themselves (gestalt systems). I am interested in searching for improvements and advances in decision-making, decision-theory, and interactive intelligent systems that allow for coping with real-time and real-world environments, and investigating the development of human-consistent systems that become more intuitive for humans to work with and feel more natural to interact with in our increasingly technological society.

    My preference is to work in applied AI, and my strength is in integrated intelligent systems—applying techniques in new and novel combinations, expanding their capability under new constraints posed by their application, and making a tangible improvement in the applied domain. I feel that my strength is in applying theoretical techniques to real-world/real-time problems.
    My current primary work is in the exploration of agent learning, faster agent creation techniques, generation of knowledge from geometry, middleware support for AI, and the integration of academic AI techniques into current electronic entertainment games.

Copyright ©2008-9 G. Michael Youngblood, Ph.D. ALL RIGHTS RESERVED.