Z. Majercik, Sharon Zhang, William Wang et al. (9 authors) • Semantic Scholar
Finding: A system using constrained solvability queries and AI (GoExplore + RL) to give game designers rapid feedback on whether and how players can solve 3D obstacle courses, including identifying unintended solution paths and difficulty assessment.
Why it matters: Directly relevant to game design tools and AI-assisted creativity/design workflows — the system functions as an AI design collaborator that expands designers' solution space awareness and supports iterative design thinking.
Method: GPU-accelerated simulation running at ~14,000× real-time combined with reinforcement learning-trained agent guides exploratory search; validated through design studies and human playtesting.
Jargon: *GoExplore* — an algorithm for exploratory search that avoids getting stuck by returning to previously visited states; *constrained solvability queries* — asking the system to find solutions that meet specific designer-defined conditions (e.g., avoid a zone, limit jump count).
Yi-Ling Huang, Mei-La Su, Ying-Ying Chen et al. (4 authors) • Semantic Scholar
Finding: Telerehabilitation adherence in stroke survivors is shaped by capability, opportunity, and motivation factors, with vocational goal-setting, serious games, and family support identified as key facilitators.
Why it matters: Tangentially relevant through the mentions of goal-setting, serious games, and behavior change frameworks (COM-B), but the core focus is clinical rehabilitation medicine, not HCI, creativity, or team dynamics.
Finding: DHH players in China engage with mainstream games for relaxation and social connection, but face accessibility barriers; gaming preferences vary by communication method (sign/written vs. spoken language users).
Why it matters: Marginally relevant to game design and accessibility, but focused narrowly on a specific disability population in China with limited connection to the professor's core interests in creativity, teams, HCI interaction design, or behavioral science.
⚠️ Self-reported data only | No control group mentioned | Single study (not replicated) | Limited statistical detail in abstract
Finding: Riddle-based learning improved cognitive, behavioral, and emotional engagement and conceptual understanding of chemistry concepts in secondary school students, with medium effect sizes.
Why it matters: Tangentially relevant to gamification/serious games and engagement mechanics in learning contexts, but the domain (secondary chemistry) and method (riddles as pedagogy) are too narrow and applied to be directly useful for the professor's courses.
Ling Cheng, Li Jiao, Shun-Yi Hu • Semantic Scholar
Finding: A generative AI framework for personalized English learning pathways outperforms traditional approaches on satisfaction, progress, and feedback latency, with highly personalized configurations showing strongest gains.
Why it matters: Tangentially touches on AI-human collaboration in learning contexts, but is narrowly focused on English language education rather than the professor's core interests in HCI, creativity, teams, or behavioral science.
Method: Three-group comparison of traditional vs. two AI-adaptive learning pathway configurations.