Multi-Class Emotion Detection (MCED) using Textual Analysis

Abstract

Stress and mood disorder is becoming a routine life illness for any human being and it is necessary to find technology based solutions for finding cure and self-treatment of such disorders. In order to find treatment and remedy, it is important to detect ones’ emotion before applying mitigation technique. Emotion plays an important role in social interaction and has strong connection with human body and brain signals. Emotions can be stated in many ways like facial expression and body language, speech and by text. Proposed technique is targeting social media platformsfor such purpose. As huge textual information is available on social media platforms such as Facebook, Twitter, YouTube etc. in the form of comments, posts etc. Emotion Detection using text is basically a content – based classification problem, connecting ideas from the areas of Natural Language Processing as well as Machine Learning. In this paper we proposed a novel way to detect emotions using Naïve Bayes algorithm by collecting person’s browsing history. To find emotions we used Plutchik’s Wheel of Classification to check the where the given emotion lies.