Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Taxonomy of User Needs and Actions

Renee Shelby, Fernando Diaz, Vinodkumar Prabhakaran

TL;DR

This work introduces the Taxonomy of User Needs and Actions (TUNA), a three-level, multi-mode framework designed to describe both instrumental goals and the social, contextual work that unfolds in AI dialogues. Grounded in 1193 public conversations and a rigorous iterative methodology, TUNA enumerates six interaction modes and 57 request types organized as modes > strategies > types, enabling multi-scale evaluation and policy alignment. By foregrounding user agency and appropriation practices, the framework clarifies how information seeking, processing, procedural tasks, content creation, social interaction, and meta-conversation co-occur and shape outcomes, with demonstrated validation across diverse dialogues. The authors argue that TUNA provides a foundational vocabulary for safer, more responsive, and more accountable conversational systems, and offers both theoretical insight and practical design guidance for cross-domain taxonomies, interface design, and governance.

Abstract

The growing ubiquity of conversational AI highlights the need for frameworks that capture not only users' instrumental goals but also the situated, adaptive, and social practices through which they achieve them. Existing taxonomies of conversational behavior either overgeneralize, remain domain-specific, or reduce interactions to narrow dialogue functions. To address this gap, we introduce the Taxonomy of User Needs and Actions (TUNA), an empirically grounded framework developed through iterative qualitative analysis of 1193 human-AI conversations, supplemented by theoretical review and validation across diverse contexts. TUNA organizes user actions into a three-level hierarchy encompassing behaviors associated with information seeking, synthesis, procedural guidance, content creation, social interaction, and meta-conversation. By centering user agency and appropriation practices, TUNA enables multi-scale evaluation, supports policy harmonization across products, and provides a backbone for layering domain-specific taxonomies. This work contributes a systematic vocabulary for describing AI use, advancing both scholarly understanding and practical design of safer, more responsive, and more accountable conversational systems.

Taxonomy of User Needs and Actions

TL;DR

This work introduces the Taxonomy of User Needs and Actions (TUNA), a three-level, multi-mode framework designed to describe both instrumental goals and the social, contextual work that unfolds in AI dialogues. Grounded in 1193 public conversations and a rigorous iterative methodology, TUNA enumerates six interaction modes and 57 request types organized as modes > strategies > types, enabling multi-scale evaluation and policy alignment. By foregrounding user agency and appropriation practices, the framework clarifies how information seeking, processing, procedural tasks, content creation, social interaction, and meta-conversation co-occur and shape outcomes, with demonstrated validation across diverse dialogues. The authors argue that TUNA provides a foundational vocabulary for safer, more responsive, and more accountable conversational systems, and offers both theoretical insight and practical design guidance for cross-domain taxonomies, interface design, and governance.

Abstract

The growing ubiquity of conversational AI highlights the need for frameworks that capture not only users' instrumental goals but also the situated, adaptive, and social practices through which they achieve them. Existing taxonomies of conversational behavior either overgeneralize, remain domain-specific, or reduce interactions to narrow dialogue functions. To address this gap, we introduce the Taxonomy of User Needs and Actions (TUNA), an empirically grounded framework developed through iterative qualitative analysis of 1193 human-AI conversations, supplemented by theoretical review and validation across diverse contexts. TUNA organizes user actions into a three-level hierarchy encompassing behaviors associated with information seeking, synthesis, procedural guidance, content creation, social interaction, and meta-conversation. By centering user agency and appropriation practices, TUNA enables multi-scale evaluation, supports policy harmonization across products, and provides a backbone for layering domain-specific taxonomies. This work contributes a systematic vocabulary for describing AI use, advancing both scholarly understanding and practical design of safer, more responsive, and more accountable conversational systems.

Paper Structure

This paper contains 51 sections, 8 figures, 10 tables.

Table of Contents

  1. Introduction
  2. Toward User-Centered Taxonomies of AI Interaction
  3. Theoretical Foundations: The Situated Nature of AI Interaction
  4. The Social and Collaborative Nature of Human-AI Interaction
  5. Evolution and Limitations of Existing Taxonomies
  6. Methodology
  7. Dialogue Corpora
  8. Defining the Taxonomy’s Meta-Characteristic
  9. Establishing Ending Conditions
  10. Approach: Iterative Development
  11. Synthesis and Structuring
  12. Validating on New Dialogues
  13. Limitations
  14. Taxonomy of User Needs and Actions
  15. Information Seeking
  16. ...and 36 more sections

Figures (8)

  • Figure 1: A flowchart of the TUNA taxonomy development.
  • Figure 2: A conceptual model of the TUNA Framework. The model illustrates how core instrumental tasks (Modes 1-4) are supported by the foundational relational layer (Mode 5) and the governing contextual layer (Mode 6), emphasizing how instrumental actions are managed through social and conversational work.
  • Figure 3: TUNA annotated dialogue with Mode 1 request types
  • Figure 4: TUNA annotated dialogue with Mode 2 request types
  • Figure 5: TUNA annotated dialogue with Mode 3 request types
  • ...and 3 more figures