Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Intent Tagging: Exploring Micro-Prompting Interactions for Supporting Granular Human-GenAI Co-Creation Workflows

Frederic Gmeiner, Nicolai Marquardt, Michael Bentley, Hugo Romat, Michel Pahud, David Brown, Asta Roseway, Nikolas Martelaro, Kenneth Holstein, Ken Hinckley, Nathalie Riche

TL;DR

This work addresses the challenge of aligning GenAI-generated content with user intent in complex, non-linear creative tasks like slide deck creation. It proposes IntentTagger, a system built on granular Intent Tags that support graphical micro-prompting, enabling flexible, non-linear, and interpretable human-AI co-creation workflows. A lab study with 12 participants shows that users prefer intent-tag interactions over chat-based prompts or design galleries, citing improved control, alignment, and meta-intent elicitation, while highlighting needs for onboarding support and element locking. The paper discusses design considerations and future work to extend intent tagging to diverse content creation tasks beyond slides, aiming to balance flexibility with usability in GenAI-assisted workflows.

Abstract

Despite Generative AI (GenAI) systems' potential for enhancing content creation, users often struggle to effectively integrate GenAI into their creative workflows. Core challenges include misalignment of AI-generated content with user intentions (intent elicitation and alignment), user uncertainty around how to best communicate their intents to the AI system (prompt formulation), and insufficient flexibility of AI systems to support diverse creative workflows (workflow flexibility). Motivated by these challenges, we created IntentTagger: a system for slide creation based on the notion of Intent Tags - small, atomic conceptual units that encapsulate user intent - for exploring granular and non-linear micro-prompting interactions for Human-GenAI co-creation workflows. Our user study with 12 participants provides insights into the value of flexibly expressing intent across varying levels of ambiguity, meta-intent elicitation, and the benefits and challenges of intent tag-driven workflows. We conclude by discussing the broader implications of our findings and design considerations for GenAI-supported content creation workflows.

Intent Tagging: Exploring Micro-Prompting Interactions for Supporting Granular Human-GenAI Co-Creation Workflows

TL;DR

This work addresses the challenge of aligning GenAI-generated content with user intent in complex, non-linear creative tasks like slide deck creation. It proposes IntentTagger, a system built on granular Intent Tags that support graphical micro-prompting, enabling flexible, non-linear, and interpretable human-AI co-creation workflows. A lab study with 12 participants shows that users prefer intent-tag interactions over chat-based prompts or design galleries, citing improved control, alignment, and meta-intent elicitation, while highlighting needs for onboarding support and element locking. The paper discusses design considerations and future work to extend intent tagging to diverse content creation tasks beyond slides, aiming to balance flexibility with usability in GenAI-assisted workflows.

Abstract

Despite Generative AI (GenAI) systems' potential for enhancing content creation, users often struggle to effectively integrate GenAI into their creative workflows. Core challenges include misalignment of AI-generated content with user intentions (intent elicitation and alignment), user uncertainty around how to best communicate their intents to the AI system (prompt formulation), and insufficient flexibility of AI systems to support diverse creative workflows (workflow flexibility). Motivated by these challenges, we created IntentTagger: a system for slide creation based on the notion of Intent Tags - small, atomic conceptual units that encapsulate user intent - for exploring granular and non-linear micro-prompting interactions for Human-GenAI co-creation workflows. Our user study with 12 participants provides insights into the value of flexibly expressing intent across varying levels of ambiguity, meta-intent elicitation, and the benefits and challenges of intent tag-driven workflows. We conclude by discussing the broader implications of our findings and design considerations for GenAI-supported content creation workflows.

Paper Structure

This paper contains 51 sections, 29 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Intent Elicitation in Human-AI Interaction
  4. Systems for Supporting Prompting and Steering GenAI
  5. Non-Linear Content Creation and Iterative Design Workflows
  6. Systems for Supporting Slide Deck Creation
  7. Exploring Design Principles for GenAI-supported Rich Content Creation
  8. Design Principles
  9. IntentTagger: An Intent Tagging-based Slide Creation System
  10. Example Use Case Scenario
  11. Interface Features
  12. Intent Tags: Concept Tags, Reference Tags and Groups (DP2)
  13. Deck Steering Canvas and Slide Steering Overlay (DP1, DP3)
  14. Outline Editor (DP1)
  15. Editing Intent Tags, Tag Widgets, and System Suggestions (DP3, DP4, DP5)
  16. ...and 36 more sections

Figures (29)

  • Figure 1: Screenshots of the IntentTagger system: (Left window) A user creates a slide presentation using the Slide Deck Steering Canvas with (1) active tags in the three circular tag groups surrounded by (2) inactive system-suggested tags. (3) The generated outline and slides are displayed in the tabbed sidebar. (Right window) The user makes (scoped) adjustments to a single slide via the Slide Steering Overlay and (4) explores generated slide variations. (5) Then, local slide changes can be accepted or applied to the entire deck.
  • Figure 2:
  • Figure 3:
  • Figure 4:
  • Figure 5:
  • ...and 24 more figures