Author Information
Danielle Pompeu Noronha Pontes, State University of Amazonas, BrazilAbstract
Interface evaluation methods play a central role in the field of Human–Computer Interaction (HCI), enabling the systematic analysis of usability, accessibility, and the visual organisation of interactive systems. However, applying these methods in educational contexts often requires methodological guidance and analytical experience from students. This paper presents UXMind, an intelligent agent designed to support AI-assisted interface evaluation in user experience (UX) education. The agent, available at https://danipontes.github.io/uxmind/, integrates multiple theoretical frameworks into a single web-based platform, enabling structured evaluations and the automated generation of interpretive analyses. UXMind operationalises 47 evaluation criteria derived from four widely recognised frameworks in the HCI literature: Nielsen’s usability heuristics, Bastien and Scapin’s ergonomic criteria, the Web Content Accessibility Guidelines (WCAG 2.1), and Gestalt principles applied to visual perception. During the evaluation process, the agent perceives the evaluation environment by collecting the scores assigned to each criterion and the qualitative observations recorded by evaluators. These data are processed by a Large Language Model (LLM) that acts as the cognitive component of the agent, enabling it to reason over usability evidence, identify evaluation patterns, and produce structured interpretations. The interpretive analysis is generated through an LLM accessed via the Groq API, a high-performance inference platform designed for low-latency execution of language models. By integrating interface evaluation frameworks with generative artificial intelligence and intelligent agent capabilities, UXMind contributes a novel approach to AI-assisted UX education, enabling scalable interface inspection practices and supporting the development of critical evaluation skills in HCI training.
Paper Information
Conference: PCE2026Stream: Design
This paper is part of the PCE2026 Conference Proceedings (View)
Full Paper
View / Download the full paper in a new tab/window
To cite this article:
Pontes D. (2026) UXMind: An Intelligent Agent for AI-Assisted Interface Evaluation in UX Education ISSN: 2758-0962 The Paris Conference on Education 2026: Official Conference Proceedings (pp. 623-635) https://doi.org/10.22492/issn.2758-0962.2026.48
To link to this article: https://doi.org/10.22492/issn.2758-0962.2026.48
Comments
Powered by WP LinkPress