Real-Time Customer Emotion Analysis in E-Commerce based on Social Media Data: Insights and Opportunities

In this era of social media, it's essential for businesses to monitor their customers options and feelings regarding their services and products in a timely manner. Due to the ease of sharing opinions and feedback on social media, the customers can share their reviews about the business or a product instantly. This feedback can have a significant impact on the business's reputation and in turn on its revenue. In this regard, sentiment analysis has developed into a vital tool that companies can use to comprehend the emotional factors that influence client behavior and to aid them in making decisions that will increase customer pleasure. This work presents the use of social media data for real-time consumer emotion analysis in e-commerce. The study aims to identify the most expressed emotions and provide businesses with the ability to tailor their product and services accordingly. The employed dataset consists of 58,000 English comments that have been labelled for 27 different emotion categories. The study uses machine learning methods to categorize the emotions expressed in the comments, including convolutional neural networks and Bidirectional Encoder Representations from Transformers (BERT). The practical result of this research shows the importance of machine learning model coupled with a user interface that can provide stakeholders, such as e-commerce companies, with insights into consumer emotion as well as realtime customer sentiment about their goods and services.