Using Students’ Performance to Improve Ontologies for Intelligent E-Learning System

Abstract

The most important goal of Artificial Intelligence (AI) is to improve computer understanding and make it as close to human intelligence as possible. One of the most significant areas of application for artificial intelligence is within the E-learning system. Most of the E-learning systems lack AI tools and merely present the content materials without evaluating the students' prior learning. However, using ontologies together with semantics for various intelligent systems have showed promising results in various fields such as real estate (Yuan, Lee, Kim, & Kim, 2013), geospatial problem-solving environment (Jung, Te Sun, & Yuan, 2013), and collaborative learning (Isotani et al, 2013). In addition, some intelligent have already been commercialized (Fensel, Van Harmelen, Horrocks, McGuinness, & Patel-Schneider, 2001). It has been showed that when ontology-based methods used the essential building block of AI, semantics and ontology can drive the E-learning systems to the next phase (Bittencourt, Costa, Silva, & Soares, 2009; Gaeta, Loia, Orciuoli, & Ritrovato, 2015; Gaeta, Orciuoli, & Ritrovato, 2009; C.-C. Hsu & Ho, 2012; Kontopoulos, Vrakas, Kokkoras, Bassiliades, & Vlahavas, 2008; Leony, Parada Gelvez, Munoz-Merino, Pardo, & Kloos, 2013; Munoz Merino & Kloos, 2008).The intelligent learning systems are basically systems that make decisions on student learning as oppose to other e-learning systems where a teacher makes all the instructional decisions (Isotani et al., 2013). The system evaluates the student's prior learning, decides what to learn next, analyzes the achievement, determines their competency level, and then directs the student to the next learning objective. One of the advantages of intelligent learning systems is that it provides students with a learning course that is specifically tailored for their individual learning style. Adaptive learning implementations include learning environments that change according the students' individual learning needs, and consequently, increase meaningful learning and student achievement (Ozyurt, Ozyurt, & Baki, 2013). Besides assessing the student's prior knowledge, an intelligent system needs to track the student's comprehension during the learning process. This can be achieved by collecting test scores and attendance from the learning activities.Ontology houses a semantic map of pieces of information collected in order to represent knowledge. Ontology is an explicit way to conceptualize and represent knowledge (Gruber, 1993). In its simplest form, it consists of a set of concepts, a body of labels describing how the concepts are related, and a set of information about relationship features of the connected concepts. In the context of education, an ontology would cover all the concepts and relationships about the subject to be taught (Gultepe & Memis, 2014). A novel approach of using ontology in E-learning systems is to use technology to make instructional decisions on L