Implementation of the Pedagogical Module in the Emotionally Intelligent Tutoring System
Summary of the Doctoral Thesis
Sintija Petroviča, Riga Technical University, Latvia
Development of intelligent tutoring systems (ITSs) started almost 50 years ago with the aim of creating intelligent systems that are able to simulate entirely the behavior of human-teachers and provide individualized tutoring process for each student taking into account his/her characteristics, e.g., knowledge level, learning style, interests, etc. Despite the progress made in this direction over the years, these systems are still less capable than human-teachers regarding their abilities to implement tailored tutoring process. A major drawback of these systems is the lack of emotional intelligence because the latest research in such fields as psychology and pedagogy has shown that emotions can significantly influence people’s behavior and cognitive processes including abilities to learn. The creation of ITSs possessing emotional intelligence requires integration of solutions that are coming from a new direction of artificial intelligence called affective computing which originated in 1997. This was a start for a new generation of tutoring systems entitled affective or emotionally intelligent tutoring systems (EITSs). Most of the research in this area has been directed to the functionality improvement of the existing ITSs to endow them with the ability to determine the emotional state of a student. However, an adaptation to the recognized student’s emotions is still underdeveloped since emotions are not considered when the behavior of ITSs or the tutoring process should be changed in order to achieve higher learning efficiency and learning outcomes. Therefore, the adaptability of developed systems stays at the same level as it was before the integration of emotions. The doctoral thesis studies ITSs and provides an analysis of their components by paying particular attention to a pedagogical module. This module is directly responsible for the course of the tutoring process and its individualization. The thesis provides a detailed study of the pedagogical module, its role, tasks, functionality and knowledge required for its operation. Moreover, pedagogical strategies and their role in the adaptation of the tutoring process are examined. In addition, the impact of emotions on the learning process and their inclusion in pedagogical strategies are studied. Furthermore, a study of existing EITSs is carried out with the aim of identifying their functioning principles, methods employed for the recognition of a student’s emotions, advantages and disadvantages of these methods and shortages in the adaptation of the tutoring process to the recognized emotions. Based on the study results, methods and algorithms are proposed for the elimination of the shortages found. The particular emphasis is placed on the tutoring adaptation by offering a two-level adaptation approach. Developed methods, algorithms and approach are integrated into the EITS as part of an educational game which is implemented with an aim of assessing students’ knowledge. The thesis describes implementation details and operating principles of the system and the game and the results of their experimental evaluation. The main results of the thesis are reflected in 15 scientific papers from which nine are presented at international scientific conferences.
|Country of Publication