Principal Component Analysis on the Students' Perception of a Cognitive Assistant for Content Reinforcement in Higher Education

Abstract

The study in this chapter presents training by highly ontology-oriented tutoring host (THOTH), a cognitive assistant applied to students of higher education. It was developed to provide a reinforcement of contents, aiming to reach a high level of interactivity between users and interfaces. THOTH is based on the theoretical assumption that knowledge is organized in the form of ontologies constructed in the ORAV model in regard to the Ausubel's meaningful learning. THOTH processes the required objects of the ontology in order to facilitate the formulation of standard questions based on the attributes. After one session, students gave its perceptions in a Likert-scale questionnaire with 13 questions. A principal component analysis was performed with 35 questionnaires revealing eight different categories of grouped questions, ranging from the degree of functionality in the learning process to featuring how users were accepting the conversations. The evaluation of the categories is explained quantitatively, highlighting relationships between the elements of each category of study.