Publication Date

Fall 2024

Degree Type

Master's Project

Degree Name

Master of Science in Computer Science (MSCS)

Department

Computer Science

First Advisor

Faranak Abri

Second Advisor

Navrati Saxena

Third Advisor

Nada Attar

Keywords

Emotion Classification, Conversations, Acoustic Features, MultiModel, Contextual Dependencies, BERT, Fusion Models.

Abstract

Emotion recognition is gaining traction due to its wide range of potential applications across different fields. With the rise of social media, chat platforms, and voice assistants, there is a vast increase in data through which humans implicitly and explicitly carry emotional cues. With new algorithms being developed for understanding the nuances of human language and emotion, businesses can tailor more personalized and empathetic service. Sentiment analysis, expresses a positive, negative, or neutral viewpoint laid the foundation of Emotion classification. Emotion classification in conversations represents the most advanced stage of classification. It is also challenging due to the existence and expression of Human emotions change from person to person in daily life. Several techniques can be used to detect emotions in text and conversations, like analyzing lexical features, machine learning, and hybrid approaches. Capturing contextual and temporal dependencies can help in the accurate prediction of Emotions in conversation. This research explores emotion classification in conversations, with textual, audio, and multimodal approaches. We experiment the effectiveness of analyzing contextual dependencies in text-based emotion detection with BERT models, also extract the acoustic features, such as MFCC, Chroma and 10 others for recognizing emotions from audio data. Additionally, implement a four different fusion techniques that combines textual and audio information to enhance the accuracy of emotion classification in conversations with various Deep Learning models.

Download

Available for download on Saturday, December 20, 2025

Share

COinS