arXiv subjects

Matheus Valentim

Publications and source records attributed to Matheus Valentim.

4 recordsLinked to original sources

arXiv2026

Towards Measuring Interactive Visualization Abilities: Connecting With Existing Literacies and Assessments

How do we assess people's abilities to interact with data visualizations? The current state-of-the-art visualization literacy tests -- such as VLAT and its derivatives -- only involve the use of static visualizations. Despite advances in investigating multiple visualization abilities, we do not yet have formal methods to assess the ability of a person to interact with a data visualization effectively. In this position paper, we discuss related literacy concepts and assessments to propose and compare different approaches for assessing the abilities that people leverage to use visualizations in interactive sensemaking tasks.

cs.HC↗

arXiv2026

A Multiliteracy Model for Interactive Visualization Literacy: Definitions, Literacies, and Steps for Future Research

This paper presents a theoretical model for interactive visualization literacy to describe how people use interactive data visualizations and systems. Literacies have become an important concept in describing modern life skills, with visualization literacy generally referring to the use and interpretation of data visualizations. However, prior work on visualization literacy overlooks interaction and its associated challenges, despite it being an intrinsic aspect of using visualizations. Based on existing theoretical frameworks, we derive a two-dimensional model that combines four well-known literacies with five novel ones. We found evidence for our model through analyzing existing visualization systems as well as through observations from an exploratory study involving such systems. We conclude by outlining steps towards measuring, evaluating, designing for, and teaching interactive visualization literacy.

cs.HC↗

arXiv2025

The Plot Thickens: Quantitative Part-by-Part Exploration of MLLM Visualization Literacy

Multimodal Large Language Models (MLLMs) can interpret data visualizations, but what makes a visualization understandable to these models? Do factors like color, shape, and text influence legibility, and how does this compare to human perception? In this paper, we build on prior work to systematically assess which visualization characteristics impact MLLM interpretability. We expanded the Visualization Literacy Assessment Test (VLAT) test set from 12 to 380 visualizations by varying plot types, colors, and titles. This allowed us to statistically analyze how these features affect model performance. Our findings suggest that while color palettes have no significant impact on accuracy, plot types and the type of title significantly affect MLLM performance. We observe similar trends for model omissions. Based on these insights, we look into which plot types are beneficial for MLLMs in different tasks and propose visualization design principles that enhance MLLM readability. Additionally, we make the extended VLAT test set, VLAT ex, publicly available on https://osf.io/ermwx/ together with our supplemental material for future model testing and evaluation.

cs.HC↗

arXiv2024

Hacc-Man: An Arcade Game for Jailbreaking LLMs

The recent leaps in complexity and fluency of Large Language Models (LLMs) mean that, for the first time in human history, people can interact with computers using natural language alone. This creates monumental possibilities of automation and accessibility of computing, but also raises severe security and safety threats: When everyone can interact with LLMs, everyone can potentially break into the systems running LLMs. All it takes is creative use of language. This paper presents Hacc-Man, a game which challenges its players to "jailbreak" an LLM: subvert the LLM to output something that it is not intended to. Jailbreaking is at the intersection between creative problem solving and LLM security. The purpose of the game is threefold: 1. To heighten awareness of the risks of deploying fragile LLMs in everyday systems, 2. To heighten people's self-efficacy in interacting with LLMs, and 3. To discover the creative problem solving strategies, people deploy in this novel context.

cs.CR↗