Text Mining for Fuzzy-based Emotion Expressions

Abstract

A model of emotion representation in the form of facial expressions using text mining technique, followed by fuzzy-based mechanisms using Fuzzy Inference System (FIS) is proposed. The model classifies the emotional content of sentences from text input and expresses corresponding emotions by a facial expression. Text input is classified using Na¨ıve Bayes text classifier, while facial expression of a virtual character are controlled by Mamdani fuzzy inference system utilizing results from text classifier. This model is able to show the facial expression with admixture blending emotions. As a demonstration, examples of facial expressions with corresponding text inputs as results from the implementation of our model are shown.