Comparative Study Between Deep Learning and Traditional Machine Learning Models for Sentiment Analysis

No Thumbnail Available
Rachayya Mathapati, Chaitanya
Issue Date
MSc in Data Analytics
Dublin Business School
Items in eSource are protected by copyright. Previously published items are made available in accordance with the copyright policy of the publisher/copyright holder.
Sentiment Analysis represents a dynamic field of research within the field of text mining. It involves the computational analysis of opinions, emotions, and subjective elements present in written text. This process entails the systematic examination of sentiments conveyed through language. The study investigates the relative effectiveness of four sentiment analysis techniques: (1) traditional supervised machine learning model using logistic regression, (2) Naive Bayes,(3) Support Vector Machine, and (4) Advanced supervised deep learning model using Bidirectional Encoder Representations from Transformers (BERT). A thorough examination was conducted on a publicly accessible dataset containing 10,261 Amazon reviews focusing on musical instruments. These reviews capture customer emotions, closely linked with associated ratings. Due to its current relevance, precise interpretation of sentence context and determination of whether it expresses positive or negative sentiment is of paramount importance. The efficacy of sentiment classification evaluation was enhanced by analysing metrics including accuracy, precision, recall, and the F1 score.