An Robust Machine Learning Methods for Sentiments Analysis of Amazon Product in E-Commerce
Praveen Kumar Mannepalli, Parul Khatri, Meenakshi Patel, Sana Khan
Abstract
The rise in popularity of online platforms like social media, search engine optimisation, and email has propelled a field of Sentiment Analysis, which seeks to decipher an underlying meaning of a text, to the forefront of global discourse. In order to do sentiment analysis on Amazon product evaluations, this article presents a comparison examination of several machine learning algorithms. SVM, DT, RFC, CNN, and LSTM models will be the primary targets of their investigation. Research using sentiment analysis involves customers who are aware of how a product makes them feel. The data collection utilised is a Kaggle of product reviews from Amazon. It turns out that the LSTM ML algorithm is the most accurate of all the ones that were tried. The results show the LSTM model shows a higher accuracy of 97%, in comparison to other models. Research like this improves both the decision-making process and the online buying experience by shedding light on customer opinion.