Students learning a particular topic perform better with private tutoring than with classroom instruction. It would seem that the degree of personalized instruction plays a significant role. The problem with private tutoring is the amount of tutors necessary for providing an individualized learning experience. In a realistic setting, e. g. in a “Tutorium” at TH Köln, a human tutor helps students further understand a subject by answering questions and helping them to solve problems tailored to their level of knowledge. This requires an understanding of a student’s skill level. Additionally, the skill level varies from student to student. Furthermore, time constraints on the human tutor can also be a hindering factor. Effective tutoring can thus become difficult. Intelligent tutoring systems (ITS) address this issue of personalized learning.
Derived Research Question
- How can intelligent tutoring systems prevent motivational troughs and assess the skill level of a student in the e-Learning context ArchiLab?
- What learning strategies and learning hierarchies help a novice achieve an expert competence level?
- How could the knowledge level of students be modeled on a conceptual/technical level?
- How could the knowledge base of an ITS be modeled on a conceptual/technical level?
- How could the architecture of an ITS look like?
- Gathering of information through study of literature and scientific papers
- Extract the useful information and build a common domain language
- Identify the most promising approaches to build an ITS
- Describe how knowledge can be modeled
- Describe how the knowledge level of students can be modeled
- Describe possible architecture of an ITS
- Identify algorithms that ITS can use to measure the students’ skill levels
- Identify algorithms that ITS can use to individually instruct students
Julian Lengelsen / Jan Seidler