RESEARCH

Publications

Article

How large language models classify and semantically explain facial expressions from valence-arousal values

LINK

Large Language Models (LLMs) primarily operate through text-based inputs and outputs, yet emotion is communicated through both verbal and non-verbal cues, including facial expressions. While Vision-Language Models analyse facial expressions from images, they are resource-intensive and may depend more on linguistic priors than visual understanding. To address this, this study investigates whether LLMs can infer affective meaning from dimensions of facial expressions—Valence-Arousal (VA) values, structured numerical representations. VA values were extracted using Facechannel from images of facial expressions (from IIMI and EMOTIC datasets) and provided to three LLMs in two tasks: (1) classifying facial expressions into basic and complex emotions and (2) generating semantic descriptions of facial expressions. The results indicate that LLMs struggle to classify VA values into discrete emotion categories, particularly for emotions beyond basic polarities. However, LLMs produced semantic descriptions that align closely with human-generated interpretations, demonstrating a stronger capacity for free-text affective inference of facial expressions.

Article

Can social media predict micro constituency and macro election outcome? exploration, insights and reflections of digital political content diffusion using data science

LINK

In democratic systems, elections represent a recurring phenomenon. The prediction of electoral outcomes and the analysis of factors influencing the success of candidates and political parties constitute a dynamic and challenging area of research, engaging diverse stakeholders, including citizens, candidates, political parties, research institutions (public and private), and civil society organizations. With advancements in digital communication technologies, information-sharing tools, and social media platforms, the processes and dynamics of elections have undergone significant transformations. These changes have introduced greater complexity to the electoral system, driven by technologically empowered citizens, rapid dissemination of information, susceptibility to misinformation, and the proliferation of fake news and hate speech. Political parties and candidates increasingly leverage social media platforms such as Facebook, Twitter, YouTube, and Instagram to disseminate their manifestos, shape public opinion in their favour or against opponents, and engage voters. These efforts often involve substantial financial investments and human resources to influence electoral outcomes. This study aims to examine and predict electoral results at the micro-constituency level based on posts and interactions on social media platforms. It explores the role of IT cells in shaping voter opinions, the influence of hate speech on electoral processes, and the quantifiable engagement of voters from diverse geographic regions in ideological discourse on social media. By employing advanced data analytics, the study seeks to understand the multidimensional impact of these factors on voter behaviour and election outcomes. The predicted results are compared against actual electoral outcomes at the micro-constituency level to evaluate the predictive model’s limitations and identify areas for further refinement.

Article

Consistency in verbal expression of death row criminals: concluding life with emotions

LINK

Introduction: 

There is a lack of both data and, more significantly, computational analyses of the linguistic behavior near an individual's final moments of life. The present study is aimed to reduce human bias and save time in psycholinguistic studies by providing data-backed insights.

Methods: 

A novel machine learning based pipeline proposed, using elements of semantic similarity (BERT and transformers) and emotion extraction in collaboration, to analyze the final statements of death row inmates to understand the consistency in their verbal expression moments before their death. A new method of analysis was proposed in this study to explore the notions inherent in the statements. A large database of 466 final statements from death row inmates in Texas was utilized in this study. Manual notion analysis was validated by a computational method of notion inferencing.

Results: 

Basic emotions of Anger and Fear majorly dominated the statements, constituting 54% of the whole, while 21% of all statements were of emotional states of Happiness and Serenity.

Discussion: 

The outcomes of this study are expected to contribute to psychological analyses of humans, moments before death, and provide insights to criminology researchers to formulate better strategies of rehabilitation and debate the death penalty.

Article

Simulation of nanocarrier-based targeted delivery of an antidepressant for major depressive disorder

LINK

Although Depression or Major Depressive Disorder (MDD) affects nearly 15% of the adult population, yet, common antidepressants that work through the monoamine hypothesis of depression are not effective against all types of MDD. The other hypothesis, blocking N-methyl D-aspartate (NMDA) receptors (NMDAR) with antagonists to promote antidepressant effects, is not very useful because of the adverse side effects of the NMDA antagonists. In this study, we have proposed a method involving the specific drug delivery of an antidepressant, using nanocarriers. The introduction of nanocarrier encapsulated potent NMDAR antagonist, R-ketamine, in the target neuronal environments of varying membrane potentials of the central nervous system was studied using in-silico coarse grain simulation of hollow spherical polypyrrole nanocarriers or increasing size. The membrane potential ruptures the nanocarrier and releases the drug R-ketamine, which has a high-binding affinity for NMDAR. Such coarse grain simulations are an interesting and emerging field of study and shed preliminary light on the theoretical behaviour of nanopharmaceuticals in vivo. The results of these in silico findings need to be validated with in vivo experimentation to develop an improved antidepressant delivery nanoformulation.

Get in touch