Schemex: Discovering Design Patterns from Examples through Iterative Abstraction and Refinement
Sitong Wang, Lydia B. Chilton
TL;DR
The paper addresses the cognitive difficulty of schema induction from example-rich domains and the need for structured guidance. Schemex provides an AI-assisted workflow with three stages—clustering, abstraction, and refinement via contrasting examples—to extract actionable schemas. Through two case studies (HCI paper abstracts and multimodal news TikToks), the approach achieves high alignment with ground-truth schemas and improves the quality of AI-generated content. This work demonstrates the potential of AI as a cognitive collaborator to reduce mental load and points to more flexible workflow design and automated iteration control as future directions.
Abstract
Expertise is often built by learning from examples. This process, known as schema induction, helps us identify patterns from examples. Despite its importance, schema induction remains a challenging cognitive task. Recent advances in generative AI reasoning capabilities offer new opportunities to support schema induction through human-AI collaboration. We present Schemex, an AI-powered workflow that enhances human schema induction through three stages: clustering, abstraction, and refinement via contrasting examples. We conducted an initial evaluation of Schemex through two real-world case studies: writing abstracts for HCI papers and creating news TikToks. Qualitative analysis demonstrates the high accuracy and usefulness of the generated schemas. We also discuss future work on developing more flexible methods for workflow construction to help humans focus on high-level thinking.