NORMA eResearch @NCI Library

Machine Learning Framework for Prediction of Empathy using Eye-tracking and Speech Analysis

Badarinath, Rahul (2022) Machine Learning Framework for Prediction of Empathy using Eye-tracking and Speech Analysis. Masters thesis, Dublin, National College of Ireland.

[thumbnail of Master of Science]
PDF (Master of Science)
Download (1MB) | Preview
[thumbnail of Configuration manual]
PDF (Configuration manual)
Download (1MB) | Preview


Empathy is often described as innate ability of an individual to perceive and sensitize to the emotional feelings of another individual and present a motivation to showcase care & affection to them. It is often believed, that empathy is the most important emotion towards building a better and more sustainable society. The objectives of this research are to help develop and improvise the current recruitment methodologies in the medical and psychological domains such as Nurses and Therapists, who require to be empathetic. The traditional approach for predicting empathy is through questionnaires. This research proposes a novel approach for predicting empathy using Machine Learning, Eye-tracking, and Speech Analysis by creating a robust framework. The framework comprises of numerous features such as Heatmaps, which were generated through Eye-tracking, most frequent emotion using Speech signals, Demographic details of a participant such as gender, age, memory test score, levels of sadness before and after watching the videos. Furthermore, metrics such as blink percentage, blink mean, blink standard deviation, saccade percentage, saccade mean, saccade, average distance from both eyes, standard deviation were also derived. A combination of these metrics were fed as inputs to three Machine Learning models; Gradient Boosting, Random Forest, and Logistic Regression.The YOLOv5 model and Principal Component Analysis were used to extract multiple features and in the preparation of data respectively. Logistic regression based model outperformed the rest with an F1 score of 0.86, Gradient Boosting a close second with an F1 score of 0.57, and Random Forest with an F1 score of 0.4..

Item Type: Thesis (Masters)
Subjects: B Philosophy. Psychology. Religion > Psychology
Q Science > QA Mathematics > Electronic computers. Computer science
T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science
R Medicine > Healthcare Industry
Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning
Divisions: School of Computing > Master of Science in Data Analytics
Depositing User: Tamara Malone
Date Deposited: 18 Jan 2023 14:57
Last Modified: 06 Mar 2023 17:15

Actions (login required)

View Item View Item