ParaAdapt: Parametric User Modeling & Adaptation for Navigating Complex Interactive Domains

ParaAdapt: Parametric User Modeling & Adaptation for Navigating Complex Interactive Domains

Theme: Many interactive systems mediate complex domains—dynamic content, multi-layer workflows, and high-dimensional option spaces—where users must orient, explore, make sense, and decide. ParaAdapt frames these systems as parameterized interaction spaces. Parameters correspond to interaction qualities and properties—mappings, constraints, scaffolding, information density, feedback, pacing, and control surfaces—making adaptation an explicit design activity.

Abstract: This workshop explores parametric interaction design as a practical approach for understanding users, building interpretable user models, and enabling adaptation/personalization across interactive artefacts and interfaces (educational tools, creativity support systems, data/AI interfaces, XR, interactive installations, playful systems, and games). A core focus is on how users navigate complex domains: layered information, multi-step tasks, conceptual spaces, and large design/decision spaces.

We connect user modelling and adaption with interaction design by treating adaptation as explicit, inspectable parametric reconfiguration—tuning guidance, constraints, information density, mappings, feedback, pacing, and content variability—rather than opaque personalization. Inspired by hands-on design activities (interaction design studios, participatory design, iterative prototyping, in-the-wild evaluation; and as one motivating example, rapid game jams using the casual game creator ParaVida), we leverage parametric traces (parameter trajectories + interaction logs + design rationale) to model navigation strategies, expertise, uncertainty, and help-seeking.

The workshop also welcomes work where parametric controls govern procedural content generation (PCG) or other generative systems as important application domains. We will produce a shared taxonomy of “complexity controls”, adaptation pattern cards, and a community roadmap/whitepaper and repository. Human-centered AI perspectives—transparency, agency, accessibility—will be integrated as design constraints to guide the development of the produced material.

Workshop overview and rationale

Rationale: UMAP methods excel at user modeling and adaptation, but complexity-heavy domains create recurring challenges: interpretability, user/designer control, and evaluation when interaction spaces are open-ended. Parametric interaction design offers an explicit “control layer” for shaping how complexity is revealed and navigated, while producing structured traces that link:

design intent → parameter changes → user behavior → outcomes

Connecting domains:

• User modeling and personalization
• Interaction design practice (prototyping, critique, participatory design, in-the-wild evaluation)
• Procedural content generation and generative pipelines (PCG) as an important—yet not exclusive—application area where parametric controls shape navigability and experience
• Human-centered AI perspectives as guardrails (transparency, agency, accessibility, contestability) for adaptive behavior

Objectives

1. Design for complex-domain navigation: Identify parametric strategies for wayfinding, sensemaking, and decision support (progressive disclosure, scaffolding, guardrails that preserve agency, navigation aids).
2. Parameterization of interaction and variability: Define reusable parameterizations grounded in interaction design semantics (affordances/signifiers, mappings, constraints, feedback, information architecture, “complexity controls”), including parameters governing content variability (e.g., via PCG or curated variation).
3. Model navigation behavior and user state: Develop pipelines linking parameters + interaction traces to user models capturing navigation strategies, uncertainty, expertise, goals, and help-seeking.
4. Adaptation as legible reconfiguration: Develop adaptation methods operating through explicit parameters with explanations, user control (override/undo), and calibration; include patterns for systems where PCG parameters are part of the adaptation surface.
5. Evaluation & governance for complex adaptive domains: Establish evaluation approaches (task success + sensemaking, longitudinal protocols, counterfactual analysis) and integrate HCAI-relevant constraints (consent, fairness, accessibility, transparency) as non-negotiable design requirements.

Alignment with UMAP: ParaAdapt advances the work presented in UMAP by (i) treating navigation in complex spaces as a first-class modeling target and (ii) grounding adaptation in explicit parametric interaction design, enabling transparent and controllable personalization across domains, including systems with PCG-driven variability.

Workshop organization

Format and duration: ~2 hours, fully in-person.
Workshop type: interactive studio/lab (artefact- and method-building, emphasizing collaboration).
Anticipated participants: ~20–35.

Expected contributions:

• Position statements (1–2 pages): concepts, frameworks, early results, datasets, evaluation protocols, ethics/UX insights
• Demo/tool submissions (1–2 pages + optional video): parametric editors, configurable interface frameworks, PCG control surfaces, instrumentation pipelines, explanation/control dashboards
• Design case studies (1–2 pages): interaction design activities (participatory design, prototyping, critique, in-the-wild deployment); creative jams welcome but not required

Workshop program

Planned outputs:

1. Taxonomy of complexity controls (interaction + variability parameters, including PCG where relevant) and corresponding user-model targets
2. A set of design pattern cards
3. Post-workshop roadmap/whitepaper outline + open repository plan

Detailed schedule (2 hours)

Active engagement strategies:

• Mixed-discipline group assignment
• Structured canvases and pattern cards to ensure tangible outputs
• Live taxonomy building to converge on shared language across communities

Submission and participation process

Submission categories:

1. Position statement (1-2 pages)
2. Demo/tool (1–2 pages + optional video)
3. Design case study (1–2 pages)

Submission format:

1. PDF
2. Single Column
3. ACMART

Deadlines:

• Submission: 21.April 2026
• Review: 28.April 2026
• Cam Ready: 8.May 2026

Review process:

• Single-blind; aim for 2 reviews per submission (Program Committee)
• Criteria: relevance to parametric user modeling/adaptation for complex-domain navigation; clarity; novelty; contribution to discussion and artefact pool; consideration of human-centric design approaches; ethics

Program Committee:

• Mathias Funk (TU/e, NL)
• Denise Lengyel (UU, SV)
• Christian Steinparz (JKU, AT)
• TBA

Participation without submission: Limited attend-only seats will be available to include practitioners/students to contribute to the labs and create a critical mass.

Organizers

Swen E. Gaudl — University of Gothenburg, Sweden (main point of contact; in-person)

Interests in interaction design and interactive artefacts, with focus on design-facing tools and methods that support iterative tuning and adaptation.
Role: overall chair; facilitation lead; synthesis/whitepaper coordination.
Email: swen.gaudl@gu.se

Mark Nelson — American University, USA (in-person, intended)

Works at the intersection of AI and interactive systems, including automated design and procedural content generation, with relevance to parameterized design spaces and controllable variability.
Role: connect PCG/generative systems to parametric user modeling; benchmark/task framing; review coordination.
Email: mnelson@american.edu

Günter Wallner — JKU Linz, Austria (in-person, intended)

Expertise in user research, behavioral analytics, and visualization—supporting interpretable user models and designer-facing understanding of navigation behavior.
Role: telemetry/user-modeling methods; interpretability + visualization lead; lab-method design.
Email: guenter.wallner@jku.at

Organizers’ diversity & commitment: The chairs span three institutions in different countries and complementary expertise (interaction design, PCG/generative systems, analytics/visualization).

Promotion and dissemination

Promotion:

• Workshop website + CFP via UMAP channels
• Cross-community outreach emphasizing UMAP × interaction design, with explicit invitation to PCG/generative-content and complex-domain interface communities:
  • HCI/interaction design and accessibility networks
  • learning tech, creativity support tools, interactive media/XR
  • PCG/generative content venues and communities
  • games and playful systems communities (as one application domain among several)

Dissemination:

1. Workshop summary for UMAP proceedings (as applicable)
2. Open repository containing taxonomy, canvases/templates, and opt-in artefacts (papers, demos, dataset schemas)
3. Community whitepaper/roadmap:
   • priority research questions and benchmark tasks
   • dataset schema suggestions including parameter trajectories and optional design rationale fields
   • guidance for transparent, agency-preserving, accessibility-first adaptation (as constraints)