Appreciation of Customer Satisfaction Through Analysis Facial Expressions and Emotions Recognition

This article presents the modelling of a system that aims to predict the satisfaction of a customer through his emotions. This system must predict customer's behavior in the decision-making process. For this end, first we extract geometric features form customer's emotional faces, captured from local camera placed near the products. Then, to predict customer satisfaction, we have classified these features using adapted SVM classifier. The kinds of customer's satisfaction are satisfied, not satisfied and neutral. Our system shows a good performance, testing it on the JAFFE dataset.