E-COMMERCE PRODUCT BASED REVIEW AND NEW STRATEGY FOR SENTIMENT ANALYSIS

Abstract

Abstract: Customers nowadays routinely share their thoughts on social media about any product, brand, or experience. Analysts collect and analyse these reviews in order to learn more about the product. The beauty of social media is that it connects people from all walks of life. As a result, the analysts gathered feedback from various social media and platform sources for practically everything. Sentiment Analysis is used to forecast outcomes to obtain relevant information, such as predicting a movie's box office success and rating new products. In today's competitive world, this form of prediction helps customers look to purchase goods or services. The goal of this article is to review e-commerce websites in a text format with some special characters and symbols (emojis). In terms of context, emotion, and prior experience, every words has some meaning. These factors influence some of the features of text data that work for prediction. This paper aims to bring together previous research on text analysis with emotion-based analysis. There is also a discussion of outstanding issues and limitations of document-based sentiment analysis. This study came to a revolutionary multi-class categorization approach as its conclusion. A ternary classification is classes divided into positive, negative, and neutral based on product-based text and emoji evaluations on the Twitter social media platform.