Real time sentiment analysis of natural language using multimedia input.

Area of Science:

• Natural Language Processing
• Affective Computing
• Machine Learning

Background:

• Accurate interpretation of semantics and sentiments is crucial for understanding intended message tone.
• Automating sentiment analysis across audio, video, and text data is a key research objective.
• Machine Learning (ML) and Deep Learning (DL) offer powerful tools for analyzing vast datasets.

Purpose of the Study:

• To compare existing sentiment analysis studies.
• To develop a novel system for real-time sentiment analysis.
• To provide users with daily attitude assessments and recommendations.

Main Methods:

• Review and comparison of previous sentiment analysis research.
• Development of a new system for real-time sentiment analysis.
• Application of various ML/DL techniques including Support Vector Machines (SVMs), Bayesian Networks (BNs), Decision Trees (DTs), Convolutional Neural Networks (CNNs), and K-Means Clustering for classification.

Main Results:

• A comparative analysis of prior sentiment analysis methodologies.
• Implementation of a functional real-time sentiment analysis system.
• Demonstration of ML/DL techniques for classifying emotions from multimedia inputs.

Conclusions:

• Sentiment analysis is vital for accurate communication interpretation.
• ML and DL techniques are effective for automating sentiment analysis.
• The developed real-time system offers practical applications for personal attitude assessment and guidance.