Understanding the relationship between user intention, behavior, interaction, and experience in metaverse environments

Abstract

This thesis presents research at the intersection of Human-Computer Interaction (HCI) and the emerging paradigm of the metaverse, focusing on user intention, behavior, interaction and experience in virtual environments. Prior research in metaverse has explored environment construction, interactive techniques and user engagement, yet these investigations have largely treated user actions as direct input-output mappings, overlooking the hierarchical cognitive structures that underlie human intentions. Understanding and predicting how users interact with objects in virtual realms, as well as effectively interpreting their natural inputs to discern intended actions, remain significant challenges. The complexity lies in deciphering user behavior within virtual environments and harnessing it to accurately interpret their input intentions, particularly as user actions are driven by mental processes with indirect mapping between behavior and interaction.

In addressing this challenge, this research investigates the relationship between user intention, behavior, interaction and experience in metaverse environments through a combination of theoretical development, technological implementation and empirical validation. This research presents three core components: a conceptual framework connecting environment construction with user intention and user factors, grounded in dynamic theory of intention and activity theory, a taxonomy of user intentions categorizing them into distal, proximal and motor levels, and a five-stage cyclical model offering design considerations for metaverse development from an HCI perspective.

Three technological developments were made for experimenting with intention-aware interaction across different metaverse contexts: the MagicBook case study for validating the intention taxonomy, the IntentVR framework for generative AI-driven interaction design, and the LanternOperAR hybrid cultural gift for user experience evaluation. Four experimental studies were conducted to investigate the interpretation of user intentions (E1), intention-behavior relationships across different individuals and contexts (E2), AI-enhanced intent-driven interaction comparing IntentVR with baseline approaches (E3), and user experience in relation to user intents (E4). The initial expert-based study validated the three-level intention taxonomy with strong information quality and satisfaction ratings. Follow-up studies found that intention-behavior translation is moderated by seven control factors (i.e., knowledge, ability, resources, availability, opportunity, cooperation and unexpected situations) and that integrating generative AI into interaction design significantly reduces task completion time while improving system usability and user engagement. Building upon these, this research further investigated how metaverse prototype design affects user experience in relation to user intents, revealing age-related differences in interaction preferences and the importance of accommodating diverse user groups. This research proposes the use of intention-aware design in metaverse environments. It offers implications for the future design, evaluation and application of virtual environments across domains that necessitate understanding and responding to user intentions.

Date of Award	15 Sept 2026
Original language	English
Awarding Institution	University of Nottingham
Supervisor	Xu Sun (Supervisor), Yoke Chin Lai (Supervisor), Haibo Zhou (Supervisor) & Cheng Yao (Supervisor)