ONTOLOGY BASED EMOTION DETECTION TRAINING TOOL

Abstract

Humans are a complex, emotion-driven machinery. When we communicate, we use not only words, but we also communicate through our body. In fact, we communicate 7% through verbal communication (words), 38% through para-verbal communication (voice) and 55% through nonverbal communication (body language, facial expression). Each emotion is instantly reflected on our faces through micro expressions, which are very short changes mainly on the eyes, lips, nose, chin, eyebrows, and forehead lines. Psychologists have been preoccupied with the universality of emotions and have identified 7 basic human emotions: fear, anger, sadness, surprise, happiness, disgust and contempt. This paper presents the design and implementation of a training tool for facial expressions detection and interpretation. The application is based on an ontological representation of human emotions and micro expressions. Each emotion is associated with specific changes at facial level, visible for a few micro seconds in the eyes, chin, forehead, nose, eyebrows and mouth. The application is completed with a user interface, where users can test their competencies in detecting human emotions based on facial expressions. The user can upload a photo, analyze it and identify specific facial features and his interpretation regarding the emotions of the person in the photo. The application interrogates the ontology and extracts the emotion displayed and compares it against the user's answer, providing the user with feedback regarding his performance. The application can be integrated into e-learning systems and can help develop human interaction and emotion detection abilities through effective training. Emotion recognition training will lead to improved results in deception detection and has direct impact in fraud detection and security, forensics, psychology, as well as organizational and interpersonal communication processes.