Conference Agenda

The proliferation of new technologies, notably GenAI, alongside augmented and virtual reality, has produced a paradoxical situation in visualization practice: while offering unprecedented multimodal richness, their complexity erects new barriers to entry, thereby exacerbating inequality and undermining design's democratic promise. Although research at the micro-level of visualization techniques abounds, a critical and systematic methodology for addressing the inherent power dynamics and design justice issues from meso- and macro-perspectives remains conspicuously absent. This paper intervenes by adapting an analytical framework from Critical Discourse Analysis to embed design justice within visualization practice. We propose a four-layer framework—Culture, Context, Meaning, and Expression (CCME)—as a shared language for the design community to systematically negotiate the tensions between technical possibility, user accessibility, and social equity. This work offers a practical yet critical tool for designers to navigate the politics of discourse in their practice, fostering a more reflective, dynamic, and liberatory approach to visualization.

This study investigates how large language models (LLMs) enable new modes of collaboration in zero-waste fashion design, addressing the challenge that traditional practice relies heavily on tacit knowledge developed through iterative material experimentation. Using first-person autoethnography, the research examines how a designer collaborates with ChatGPT and Claude across concept ideation, digital pattern development, toile construction,

and sample production, tracing how technical terminology, design reasoning, and material knowledge are negotiated, misinterpreted, and re-articulated in human–AI interaction. Findings demonstrate that ChatGPT excels at generative conceptual prompting, whereas Claude better handles complex pattern-cutting rules; nevertheless, both require iterative, dialogue-based correction. The research proposes the AI+Zero-Waste Fashion Design Process Model, which provides designers with practical strategies for integrating LLMs into zero-waste design practice.

This study explores the Miyanohara participatory land art project in Ōmuta City, Japan, as an innovative teaching method for integrating international exchange students into new cultural and environmental settings within higher education institutions (HEIs). By creating a 74-metre installation depicting the Kasumi salamander at a former coal mine site, eleven international students from Kyushu University's Strategic Design course engaged in embodied learning that transcended linguistic and cultural barriers. Using arts-based methodologies and an eco-social design approach, we demonstrate how land art can serve as a powerful tool for HEIs to foster environmental awareness, cultural integration, and transformational learning experiences for international students. The project highlights how arts-based research and eco-social design can effectively address the dual challenges of sustainability education and international student integration, thereby supporting universities' broader mission as agents of sustainable change.

Language functions as both a core element and the cognitive framework that supports the entire design process. However, the deep integration of Large Language Models (LLMs) is shifting the creative process from self- or human-to-human dialogue to human-AI conversation, requiring a careful examination of this emerging form of collaborative language. This study conducted an experiment on designer-AI collaboration. The goal was to explore the relationship between their collaborative activities, linguistic features, and co-creativity. By carefully coding activities and using the Linguistic Inquiry and Word Count (LIWC) method, we identified conversational features linked to different levels of creativity. These findings not only enhance understanding of the Human-AI collaborative design language but also offer important warnings and essential regulatory insights for designers engaging in co-creation with Textual GenAI.

Smart product concept exploration unfolds in dynamic and diverse use situations (DDUS) that yield fragmented, multimodal scenario data, yet how experienced designers organise design reasoning in data-intensive environments to support problem framing remains underdescribed. We used a Scenario-Data-Informed (SDI) Canvas to make designer-data conversations visible and ran task-based sessions with 18 designers. From coded think-alouds and canvas inscriptions, we report three findings: first, three reasoning stances that structure work with scenario data; second, three language-in-use strategies and an actor-network mapping that renders translations among contexts, artefacts, and intentions as a traceable network; third, shareable cues that support collaboration and surface frictions. We position the combination of the SDI Canvas and our actor-network map as a medium for dialogic engagement with scenario data and as a carrier of shared language, sustaining data-informed concept exploration and cross-role shared understanding in smart product design.