Graduation Semester and Year
Summer 2026
Language
English
Document Type
Dissertation
Degree Name
Doctor of Philosophy in Computer Science
Department
Computer Science and Engineering
First Advisor
Dr. Cesar Torres
Abstract
Creative practice is a continuous process of learning, experimentation, and material negotiation. Yet much of what makers learn through practice remains tacit, difficult to articulate, and difficult to transfer. Although digital fabrication, online communities, and generative AI have made creative resources widely accessible, they offer limited support for understanding how craft knowledge develops, how creative processes can be documented and shared, and how practitioners can continue developing their skills over time. In this dissertation, I investigate how computational systems can advance craft inquiry, drawing on Richard Sennett's three modes of locating, questioning, and opening. Rather than replacing the exploratory nature of making, I examine how computation can make craft knowledge more explicit, reusable, and supportive of continued learning. I develop three complementary design probes. Kilnforms explores how design spaces can be structured to support material exploration. TuneCatalog investigates how craft processes and embodied tuning knowledge can be documented and represented. Proxima examines how knowledge graphs and generative AI can recommend reachable learning opportunities based on a maker's existing repertoire.
Across these probes, I observe a consistent pattern, where different stages of craft inquiry require different forms of computational support. My findings suggest that locating benefits from structured design spaces, questioning from documented experiential knowledge, and opening from recommendations calibrated to what a maker can reach next. Through these probes, I contribute methods for representing design spaces, documenting embodied craft knowledge, and operationalizing personalized learning recommendations using knowledge graphs and generative AI. These findings suggest design implications for AI-assisted material practice by emphasizing exploration, reflection, and skill development aligned with a maker's repertoire and stage of inquiry. They offer an alternative to automation-centered approaches by demonstrating how computation can complement, rather than replace, opportunities for learning and creative growth.
Keywords
human-computer interaction, creativity support tools, retrieval-augmented generation, knowledge graphs, learning recommendation, creative exploration
Disciplines
Human-Computer Interaction
License
This work is licensed under a Creative Commons Attribution 4.0 International License.
Recommended Citation
Rakib, Mohammad Abu Nasir, "Structuring Craft Practice: An AI-Assisted Framework for Representing, Documenting, and Recommending Material Craft Knowledge" (2026). Computer Science and Engineering Dissertations. 22.
https://mavmatrix.uta.edu/cse_dissertations2/22