Sentiment Analysis Using Natural Language Processing on Text Messages

Abstract

In this digitalized age, e-commerce is rising popularity since it gives products to consumers' doorsteps without them having to leave their residences. The value of reviews has increased as buyers rely on them to make educated purchases. Machine learning may assist the task of sorting through hundreds of reviews by categorising and learning from them. Sentiment analysis, which focuses on understanding emotions and attitudes, is a key application of Natural Language Processing (NLP). This study focuses on analyzation on sentiments of Amazon product ratings utilising methods for supervised learning. The dataset utilised contains thousands of reviews across different categories. Various NLP models, including Recurrent Neural Networks (RNNs), are experimented with for categorising reviews into good, negative, or neutral attitudes. The way in which the models is tested utilising recall, accuracy, and precision. The ramifications of sentiment analysis results for businesses and customers on the Amazon platform are being investigated. This research delivers insights into sentiment analysis on Amazon's dataset and its practical applications. E-commerce is become increasingly popular, and sentiment analysis applying machine learning may aid analyse vast amounts of data to recognise emotions efficiently. Various machine learning methods, including K-Means Clustering, Decision Trees (DTs), Convolutional Neural Networks (CNNs), Support Vector Machines (SVMs), and Bayesian Networks (BNs), have been employed for emotion categorization. This research contains a comparison of earlier studies and provides a real-time sentiment analysis system to track everyday sentiments and give acceptable suggestions to users. Sentiment analysis is a crucial aspect of Natural Language Processing (NLP) that focuses on categorising sentiment polarity. This research seeks to solve issues in sentiment analysis and presents a general approach for categorising sentiment in online product reviews from Amazon.com. Sentence-level and review-level classification investigations are done, providing positive results. Future sentiment analysis work is also mentioned.