A Multi-Model Approach to Aspect-Based Sentiment Analysis of Car Reviews

This study presents an application of Aspect-Based Sentiment Analysis (ABSA) on a dataset of car reviews collected from the CarDekho website. The dataset comprises reviews for four distinct car models of four popular brands. Our objective was to extract and analyze sentiments related to ten different aspects of the cars, including performance, build quality, price, etc. We employed five different machine learning and deep learning models, namely Logistic Regression, Naive Bayes, Support Vector Machine (SVM), a simple Neural Network (NN), and the Recurrent Neural Network-Long Short-Term Memory (RNN-LSTM). Each model was trained and tested separately on the split dataset for each aspect, and its performance was evaluated based on its ability to accurately classify the sentiment of the reviews at an aspect level. The results of this study provide insights into the effectiveness of different models in performing ABSA on car reviews. This could potentially assist both consumers in making informed purchasing decisions and manufacturers in gaining a nuanced understanding of customer feedback.