Emotion reflective entity (EREN): Teaching emotional intelligence to children through storytelling with a conversational agent

• Main Author: Santos, Kyle-Althea Francesca Anne M.
• Format: text
• Language: English
• Published: Animo Repository 2019
• Subjects: Emotional intelligence; Emotions in children; Storytelling; Intelligent agents (Computer software)
• Online Access: https://animorepository.dlsu.edu.ph/etd_masteral/6527; https://animorepository.dlsu.edu.ph/context/etd_masteral/article/13521/viewcontent/Final_Document___EREN_Diary_Chatbot_for_Children2.pdf
• Institution: De La Salle University

Emotions create feelings that have effects on someone's behavior while emotional intelligence (EI) is the ability to control emotions. The best time to teach EI is from 5 to 10 years old through parental guidance; however, not all children have a parent figure and can hinder the emotional growth. This thesis suggests leveraging technology in the form of a conversational agent, Emotional Reflective ENtity (EREN), in teaching EI to children through storytelling. EREN is an empathetic conversational agent that can detect emotions from the child's utterance, and reply accordingly. Utilizing a dialogue model that is patterned after Gottman, Katz, and Hooven (1996)'s Emotion Coaching Approach, EREN helps the child sort out his/her emotions, while listening and empathizing as the child narrates his/her story. To do this, EREN adapted Oral Storytelling ENtity (ORSEN)'s Feedback, Pumping and Recollecting dialogue moves, and introduced Confirming and Recollecting dialogue strategies. Results from validation with a psychologist and end users showed that EREN is able to detect emotions correctly with an accuracy of 80.56% and 71.39% respectively. Child participants, on the other hand, acknowledged EREN as a human (3%), a friend (3%), and a therapist (3%).