2014
_NEWS Artificial Intelligence

What’s in a Game?
An interdisciplinary team is trying to better understand how to facilitate game-based learning using artificial intelligence and virtual reality.

_Jichen Zhu

Zhu is an assistant professor of digital media in the Westphal College of Media Arts & Design.

_Glen Muschio

Muschio is an associate professor of digital media in the Westphal College of Media Arts & Design.

When you’re playing a game, what does your playing style say about the way you learn? Do you just want to figure out what it takes to win, or do you prefer to fully explore the game and master the techniques?

A cross-disciplinary team of Drexel professors, with the help of a $150,000 National Science Foundation grant, is trying to improve students’ learning success by making games that can automatically adapt to a student’s playing style: a person’s tendency to be either a goal-seeker or an explorer. The project uses artificial intelligence (AI) to help kids to recognize their strengths and weaknesses as learners, then to foster the self-awareness needed to evaluate new situations.

“The idea is to use AI to help self-regulate learning,” says School of Education professor Aroutis Foster, who is collaborating with digital media professors Jichen Zhu and Glen Muschio. The game is based in part on Foster’s PCaRD–play-based pedagogical model, which advances learning through a synergistic process including play, curricular activity, reflection and discussion in iterative cycles.

The immersive learning environment is an outgrowth of Muschio’s interest in cultural heritage, and is based on Charles Willson Peale’s early-19th-century art and natural history museum housed for a time in Philadelphia’s Independence Hall. Peale, best known for his portraits of George Washington and other founders of the nation, also collected botanical and biological specimens. The virtual Peale museum serves as the setting for the game, during which the player explores the galleries and “earns” fossilized bones by solving problems, eventually assembling the fossils into the complete skeleton of the extinct mastodon.

In the learning environment, “certain options become available or not available,” explains Muschio. “But the player is unaware that their choices are being guided to the sweet spot between exploratory and goal-directed behavior.”

Zhu’s research on the state-of-the-art AI technique called “experience management” allows the game to adjust based on how the player interacts to “make the game more adaptive, more personalized,” says Zhu. The game records a player’s every move, then extrapolates from that data to change the learning environment. While the player maintains autonomy and freedom, the artificial intelligence creates subtle guidance. Zhu likens it to a movie director, overseeing and modifying the experience.

Both goal-seekers and explorers can ultimately be successful learners. Goal-seekers, however, often miss the nuances while focused on getting the grade, whereas explorers tend to value the knowledge itself. The hope is to allow students to nurture their innate tendency while also nudging them to tap into the flip side — to get an explorer, for example, to know how to focus on a goal when needed, such as working fast and finishing an assignment, and to get a goal-seeker to master subjects for the long-term. Raising one’s awareness about how he or she learns will produce more efficient learners.

“The better self-regulated you are,” says Foster, “the better learner you will be, the better student you will be, and the better you will be prepared for workplace conditions.”

Solve Problems, Earn Bones

SOLVE_PROBLEMS,_EARN_BONES

A rendering of the interactive game version of the Peale Gallery that showcases the Great American Incognitum (mastodon) skeleton. The real-life mastodon was excavated by one of Charles Willson Peale’s scientific expeditions. The player uncovers the Great Incognitum bones by solving different problems and assembles the entire skeleton at the end of the game.