We present a topological pipeline for automated multiclass emotion recognition from eye-tracking data. Delay embeddings of gaze trajectories are analyzed using persistent homology. From the resulting persistence diagrams, we extract shape-based features such as mean persistence, maximum persistence, and entropy. A random forest classifier trained on these features achieves up to $75.6\%$ accuracy on four emotion classes, which are the quadrants the Circumplex Model of Affect. The results demonstrate that persistence diagram geometry effectively encodes discriminative gaze dynamics, suggesting a promising topological approach for affective computing and human behavior analysis.

This study aimed to develop a system that provides vibrotactile feedback corresponding to the emotional content of text when a communication robot speaks. We used OpenAI's "GPT-4o Mini" for emotion estimation, extracting valence and arousal values from the text. The amplitude and frequency of vibrotactile stimulation using sine waves were controlled on the basis of estimated emotional values. We assembled a palm-sized tactile display to present these vibrotactile stimuli. In the experiment, participants listened to the robot's speech while holding the device and then evaluated their psychological state. The results suggested that the communication accompanied by the vibrotactile feedback could influence psychological states and intimacy levels.

Previous active inference accounts of emotion translate fluctuations in free energy to a sense of emotion, mainly focusing on valence. However, in affective science, emotions are often represented as multi-dimensional. In this paper, we propose to adopt a Circumplex Model of emotion by mapping emotions into a two-dimensional spectrum of valence and arousal. We show how one can derive a valence and arousal signal from an agent's expected free energy, relating arousal to the entropy of posterior beliefs and valence to utility less expected utility. Under this formulation, we simulate artificial agents engaged in a search task. We show that the manipulation of priors and object presence results in commonsense variability in emotional states.

In recent years, the fields of Music Information Retrieval (MIR) and Music Emotion Recognition (MER) have received significant attention, leading to multiple advances in how music is analyzed [1, 2]. These developments have increased the accuracy in determining what emotions are present in a given music sample, but the current state of the art is only now passing 75% through the use of Random Forest and Support Vector Machine models [3]. This is in contrast to the field of speech recognition, where current models are approaching 100% accuracy across hundreds of languages for word identification [4] and 85% for standard speech emotion recognition [5]. The additional challenges in music recognition come from the nature of music itself as the lyrical and emotional content of a vocalist's contribution are only one part of the whole. Tempo, rhythm, timbre, instrumentation choice, perceived genre, and other factors contribute together to shape the emotional and tonal landscape of any given work into a unique blend that is interpreted subjectively by individual listeners [6]. The goal of our paper is to show that by changing the underlying structure of a small subset of musical features of any given musical piece, we can adjust the perceived emotional content of the work towards a specific desired emotion.

Sentiment or mood can express themselves on various levels in music. In automatic analysis, the actual audio data is usually analyzed, but the lyrics can also play a crucial role in the perception of moods. We first evaluate various models for sentiment analysis based on lyrics and audio separately. The corresponding approaches already show satisfactory results, but they also exhibit weaknesses, the causes of which we examine in more detail. Furthermore, different approaches to combining the audio and lyrics results are proposed and evaluated. Considering both modalities generally leads to improved performance. We investigate misclassifications and (also intentional) contradictions between audio and lyrics sentiment more closely, and identify possible causes. Finally, we address fundamental problems in this research area, such as high subjectivity, lack of data, and inconsistency in emotion taxonomies.

Sentiment analysis is a continuously explored area of text processing that deals with the computational analysis of opinions, sentiments, and subjectivity of text. However, this idea is not limited to text and speech, in fact, it could be applied to other modalities. In reality, humans do not express themselves in text as deeply as they do in music. The ability of a computational model to interpret musical emotions is largely unexplored and could have implications and uses in therapy and musical queuing. In this paper, two individual tasks are addressed. This study seeks to (1) predict the emotion of a musical clip over time and (2) determine the next emotion value after the music in a time series to ensure seamless transitions. Utilizing data from the Emotions in Music Database, which contains clips of songs selected from the Free Music Archive annotated with levels of valence and arousal as reported on Russel's circumplex model of affect by multiple volunteers, models are trained for both tasks. Overall, the performance of these models reflected that they were able to perform the tasks they were designed for effectively and accurately.

Fluent human-robot collaboration requires a robot teammate to understand, learn, and adapt to the human's psycho-physiological state. Such collaborations require a computing system that monitors human physiological signals during human-robot collaboration (HRC) to quantitatively estimate a human's level of comfort, which we have termed in this research as comfortability index (CI) and uncomfortability index (unCI). Subjective metrics (surprise, anxiety, boredom, calmness, and comfortability) and physiological signals were collected during a human-robot collaboration experiment that varied robot behavior. The emotion circumplex model is adapted to calculate the CI from the participant's quantitative data as well as physiological data. To estimate CI/unCI from physiological signals, time features were extracted from electrocardiogram (ECG), galvanic skin response (GSR), and pupillometry signals. In this research, we successfully adapt the circumplex model to find the location (axis) of 'comfortability' and 'uncomfortability' on the circumplex model, and its location match with the closest emotions on the circumplex model. Finally, the study showed that the proposed approach can estimate human comfortability/uncomfortability from physiological signals.

Large language models are becoming increasingly integrated into our lives. Hence, it is important to understand the biases present in their outputs in order to avoid perpetuating harmful stereotypes, which originate in our own flawed ways of thinking. This challenge requires developing new benchmarks and methods for quantifying affective and semantic bias, keeping in mind that LLMs act as psycho-social mirrors that reflect the views and tendencies that are prevalent in society. One such tendency that has harmful negative effects is the global phenomenon of anxiety toward math and STEM subjects. Here, we investigate perceptions of math and STEM fields provided by cutting-edge language models, namely GPT-3, Chat-GPT, and GPT-4, by applying an approach from network science and cognitive psychology. Specifically, we use behavioral forma mentis networks (BFMNs) to understand how these LLMs frame math and STEM disciplines in relation to other concepts. We use data obtained by probing the three LLMs in a language generation task that has previously been applied to humans. Our findings indicate that LLMs have an overall negative perception of math and STEM fields, with math being perceived most negatively. We observe significant differences across the three LLMs. We observe that newer versions (i.e. GPT-4) produce richer, more complex perceptions as well as less negative perceptions compared to older versions and N=159 high-school students. These findings suggest that advances in the architecture of LLMs may lead to increasingly less biased models that could even perhaps someday aid in reducing harmful stereotypes in society rather than perpetuating them.

We define `ousiometrics' to be the study of essential meaning in whatever context that meaningful signals are communicated, and `telegnomics' as the study of remotely sensed knowledge. From work emerging through the middle of the 20th century, the essence of meaning has become generally accepted as being well captured by the three orthogonal dimensions of evaluation, potency, and activation (EPA). By re-examining first types and then tokens for the English language, and through the use of automatically annotated histograms -- `ousiograms' -- we find here that: 1. The essence of meaning conveyed by words is instead best described by a compass-like power-danger (PD) framework, and 2. Analysis of a disparate collection of large-scale English language corpora -- literature, news, Wikipedia, talk radio, and social media -- shows that natural language exhibits a systematic bias toward safe, low danger words -- a reinterpretation of the Pollyanna principle's positivity bias for written expression. To help justify our choice of dimension names and to help address the problems with representing observed ousiometric dimensions by bipolar adjective pairs, we introduce and explore `synousionyms' and `antousionyms' -- ousiometric counterparts of synonyms and antonyms. We further show that the PD framework revises the circumplex model of affect as a more general model of state of mind. Finally, we use our findings to construct and test a prototype `ousiometer', a telegnomic instrument that measures ousiometric time series for temporal corpora. We contend that our power-danger ousiometric framework provides a complement for entropy-based measurements, and may be of value for the study of a wide variety of communication across biological and artificial life.

Most current affect scales and sentiment analysis on written text focus on quantifying valence (sentiment) -- the most primary dimension of emotion. However, emotions are broader and more complex than valence. Distinguishing negative emotions of similar valence could be important in contexts such as mental health. This project proposes a semi-supervised machine learning model (DASentimental) to extract depression, anxiety and stress from written text. First, we trained the model to spot how sequences of recalled emotion words by $N=200$ individuals correlated with their responses to the Depression Anxiety Stress Scale (DASS-21). Within the framework of cognitive network science, we model every list of recalled emotions as a walk over a networked mental representation of semantic memory, with emotions connected according to free associations in people's memory. Among several tested machine learning approaches, we find that a multilayer perceptron neural network trained on word sequences and semantic network distances can achieve state-of-art, cross-validated predictions for depression ($R = 0.7$), anxiety ($R = 0.44$) and stress ($R = 0.52$). Though limited by sample size, this first-of-its-kind approach enables quantitative explorations of key semantic dimensions behind DAS levels. We find that semantic distances between recalled emotions and the dyad "sad-happy" are crucial features for estimating depression levels but are less important for anxiety and stress. We also find that semantic distance of recalls from "fear" can boost the prediction of anxiety but it becomes redundant when the "sad-happy" dyad is considered. Adopting DASentimental as a semi-supervised learning tool to estimate DAS in text, we apply it to a dataset of 142 suicide notes. We conclude by discussing key directions for future research enabled by artificial intelligence detecting stress, anxiety and depression.