True language understanding for an explainable AI system

• Main Author: Farhan Khalifa Ibrahim
• Other Authors: Li Fang
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2022
• Subjects: Engineering::Computer science and engineering
• Online Access: https://hdl.handle.net/10356/157406
• Institution: Nanyang Technological University

Millions of messages and thousands of articles are posted every day, and this information is stored in an unstructured natural text. Natural Language Processing (NLP) is a study to understand text using computational techniques. One of the most important tasks in NLP is sentiment analysis which studies people’s opinions, emotions, and attitudes. Sentiment analysis is a challenging task involving context understanding, language use, and unstructured human text. This project aims to use sentiment analysis techniques using different deep learning techniques. It will focus on binary sentiment classification, which detects the polarity in a text into 2 classes, positive and negative. This project studied different sentiment analysis techniques such as VADER,SVM, Naïve Bayes CNN,RNN, LSTM, GRU, and BERT. BERT gives the best accuracy among the available techniques but with the drawback that it takes a longer time to train.