Affective Interaction based Hybrid Approach for Emotion Detection using Machine Learning

Affective communication is a fresh region of research interest emerging in the interaction designers. This research will explore our hybrid approach's potential depends on both, Lexical and machine learning methods to detect the six emotional categories of Ekman in the text of the user. Initial results of the presentation evaluation of the suggested hybrid method is promoting and similar to relate the research. A demonstrative mobile application that uses the suggested strategy has been created to engage the users in a dialog that requests their views on multiple daily occurrences and offers appropriate affective answers. The novelty of this strategy is the selection of characteristics and the generation for classification of a single feature vector. Our primary objective is to improve the precision of voice classification of emotions by taking into account audio characteristics. Standard techniques like Natural Language Processing, Support Vector Machines, WordNet Affect and SentiWordNet are used in this job.