Customer Satisfaction Analysis based on Online Products by their Emotion Recognition using Meta Heuristic MachineLearning Algorithms

Abstract

In the modern corporate world, customer satisfaction is crucial for establishments. There are manual techniques, such as surveying clients and giving them questionnaires. On the other hand, companies and marketers are searching for rapid methods to obtain useful and efficient feedback for their prospective clients. Analysing evaluations by hand in order to make choices and go over business models is difficult. Natural language processing (NLP) allows users to analyse and automate this process, though.This research proposes novel technique in online product-based customer satisfaction analysis by emotion recognition using meta-heuristic machine learning (ML) model. Input is collected as online product-based customer reviews and processed for noise removal, normalization and missing value removal. Then this data feature extraction is carried out using convolutional regressive LSTM Gaussian model and classified using particle transfer graph swarm wolf optimization model. the classified output shows the detected emotion of the customer review and their satisfaction-based analysis. Experimental analysis is carried out in terms of average accuracy, mean precision, recall, F-1 score, RMSE.The proposed technique average accuracy 99%, mean precision 96%, recall 95%, F-1 score 94%, RMSE 58%.