Towards Universal Visualisation of Emotional States for Information Systems

Michal R Wrobel, Agnieszka Landowska, Karolina Makuch · 2026

Read Towards Universal Visualisation of Emotional States for Information Systems by Michal R Wrobel, Agnieszka Landowska, Karolina Makuch (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-10756.txt.

Introduction

Wrobel, Landowska, and Makuch study how people prefer emotions visualized in information systems, testing color, size, speed, shape, and animation type. Asking 419 participants about their preferences, they find color, speed, and size track selected discrete emotion labels, while speed also tracks arousal in a dimensional model. The work is a first step toward a universal emotion representation for information systems, linking visual choices to both discrete and dimensional emotion models.

Catalog record

CREATORMichal R Wrobel, Agnieszka Landowska, Karolina Makuch
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE33210
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2604.10756

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Towards Universal Visualisation of Emotional States for Information Systems Thanks: This is the postprint of a paper presented at the 32nd International Conference on Information Systems Development (ISD 2024) and published in its proceedings. DOI: 10.62036/ISD.2024.117 Michal R Wrobel, Agnieszka Landowska, Karolina Makuch Abstract The paper concerns affective information systems that represent and visualize human emotional states. The goal of the study was to find typical representations of discrete and dimensional emotion models in terms of color, size, speed, shape, and animation type. A total of 419 participants were asked about their preferences for emotion visualization. We found that color, speed, and size correlated with selected discrete emotion labels, while speed correlated with arousal in a dimensional model. This study is a first step towards defining a universal emotion representation for use in information systems. Keywords: affective computing, emotion visualisation, affective information systems 1 Introduction Emotional states are among the factors that moderate our engagement and focus. Affect influences the way people interact with information systems and the user experience. Automatic emotion recognition solutions have found applications in diverse fields [ 1 ] such as opinion mining, market research, therapy, but also in communication and information systems. The affective computing domain has developed not only the technologies to capture human affect, but also the solutions to express and influence affect. When an emotional state is roughly quantified in a software environment, the result of the analysis should be effectively communicated to a potential human user to close the affective loop. Representations of emotional states should be easily perceived and understood by humans. There are a number of ways to communicate an emotional state within an information system, including labels, numbers, and visual representations. Visualization of emotional states has received some research attention in the last 10 years. A literature review conducted in 2022 by J. Wang et al. identified a number of approaches to visual encoding, although the majority of studies focused on visualizing emotions derived from music [ 2 ] . Visual representations can include avatar facial expressions, emojis, graphs/charts, graphs with labels, emojis or icons, colors, and shapes with/without animation. In this paper, we focus on the visual representation of affect within information systems in general, specifically exploring graphical and animation-based options available in an interface, assuming that few information systems involve a facial-expressive avatar. For example, a remote communication system proposed in a study by Ertay et al. uses emoji and colors as emotion representation, without exploration of color assignment to a particular affective state (happy is yellow, fearful is pink, and disgusted is green, surprise is black) [ 3 ] . Another system for patient-physician communication uses colorful charts enhanced with emojis and labels (fear is yellow, sad is orange, surprise is turquoise) [ 4 ] . A system for visualization of emotions in calligraphy uses colors only (calm is orange, anxious is aubergine, relaxed is green) [ 5 ] . Another interesting study uses shape (cross-to-circle conversions), brightness, and size to represent pleasure-arousal-affinity dimensions of an emotional state [ 6 ] . While in the study the usage of X-cross and circle is justified by human hand gestures (body language), the choice of the other representation methods is not discussed. However, the methods proposed in research papers for representing and visualizing emotional states in information systems are often contradictory, which shows the research gap that this article aims to fill. The aim of the study was to find typical techniques for the visual representation of emotions that could be interpreted in the same way by users of an information system. Both discrete and multidimensional models of emotions were considered. The research question addressed in this paper could be formulated as follows: Is it possible to define a typical representation for the visualization of a specific emotional state in terms of color, speed, size, shape, and animation ? The paper describes a questionnaire-based study about visual representation of particular emotions and is organized as follows. The related works section provides a review of our starting point – the emotion models, and emotion visualization techniques. The research methodology section describes the concept of the study, survey design decisions, and definition of variables. Then the results section provides findings on colors, size, speed, and animation types used in emotion visualization and is followed by discussion and conclusions sections. 2 Related Work The studies that are most relevant to our research are those that deal with the representation and modeling of emotional states and those that aim to find a visualization for emotional states. Two approaches to representing emotional states are commonly used: discrete (based on labels) and multidimensional [ 7 ] . The first of this group, such as Ekman’s six basic emotions [ 8 ] or the PANAS model [ 9 ] , are more easily understood by humans and are therefore often used in surveys. On the other hand, multidimensional models, such as the Circumplex Model of Affect [ 10 ] or the VAD (valence-arousal-dominance) [ 11 ] , are more amenable to computer processing, and their mathematical representation allows them to be processed by algorithms for analyzing or synthesizing emotions. In emotion visualization, an emotional state described using the above models is mapped into a specific color, shape, animation, etc. to provide a sensory experience to a receiver of the state. The visualization aims at communication or enhancement of emotion-related information. In the study [ 12 ] authors proposed a concept of emotion-prints as an approach to visualize user emotional valence and arousal. They focused on real-time visualization of emotions to be used in standard GUI interfaces and formulated a thesis that “the representation needs to be as intuitive as possible” so that users could perceive and interpret visualization during real-time interaction. There are studies for color-based expressivity with a background in art [ 13 , 14 ] , however, those only partially transfer to information systems. There are some studies on how to visualize emotions in computer systems, but they focus on one feature (e.g. color) or one context (e.g. music). For example, so-called emotion scents based on color were proposed for valence-arousal plain representation for music-based emotions [ 15 ] . Two studies explore diverse graphs for color-based representation of emotion state change in time [ 16 , 17 ] . Animated visualizations of Self-Assessment-Manikin (a well-established questionnaire for reporting emotional states in a dimensional model) were also explored for use as a self-reporting tool [ 18 ] . An interesting study on color [ 19 ] asked people to design palettes for states such as calm, playful, exciting, and disturbing. There are a few literature reviews on emotion visualization. One study by Pinilla et.al. explores visualization for virtual reality systems [ 20 ] , highlighting object roundness, color in terms of brightness and saturation, and textures, applied to the positive-negative dimension of the emotional state model. Another study explores the visualization of sentiment from text, proposing a set of color-based tags and graphs, as well as quite interesting aggregation graphs representing anger and happiness [ 21 ] . A review study by Wang et.al. [ 2 ] examines a variety of visual encoding techniques for representing affect that have been used in

Read the complete text