Sentiment Analysis of Customer for Ecommerce by Applying AI

Conventional and manual systems are being upgraded to autonomous systems using AI techniques for better decision making and improved efficiency. Artificial Intelligence is getting more evolved and have numerous numbers of applications in the field of Health care, IOT, Security, Smart Systems, Decision makings, Intelligent computing, Digital marketing, User experience etc. In the era of digital marketing organizations face various issues regarding promotion, marketing and getting feedback from customers concerning to their products and its difficult for them to retain customer feedback in order to improve their products. In terms of estimating how much a client want to purchase this product, the offered approach gives a solution for prediction of consumer insights on the particular product. The methodology consists of two main models, the first consists of a customized Recurrent Neural Network (RNN) for user-specific perception analysis on social media and the second consists of a machine study model where classification algorithms, such as Random Forest, Naïve Bayes and BayesNet have been utilized to predict customer willingness to purchase a particular product. Furthermore, the model does not only conduct a customer sentiment analysis in relation to the product’s social media marketing in the online industry, but it is also capable of predicting if a consumer would buy it or not. The RNN model has a training accuracy of 97.15% and validation precision of 94.25% whereas the Random Forest method offers a training accuracy of 97% and 94.97% respectively.