This video tutorial explains how to perform sentiment analysis using machine learning on amazon product reviews dataset.
The dataset used for the project is obtained from Kaggle and consists of nearly 3000 reviews of amazon users regarding various amazon Alexa products like Alexa echo, Alexa dot etc. Exploratory data analysis is performed on the dataset to analyse various columns and the data is visualized using count plots and pie charts. The reviews are then processed using various methods which involve lowercase conversion, URL removal, punctuation removal, tokenisation, stop word removal and stemming. The processed data is then separated into positive and negative reviews and are then visualized using Word clouds, as word clouds help to identify the most prominent/frequently used words. Machine Learning is then performed on the processed data using various machine learning classifiers such as Logistic Regression and Multinominal Naïve Bayes.
