• The KIPS Transactions:PartA
• /
• v.10A no.5
• /
• pp.589-594
• /
• 2003

In this paper, we exmined the effect of leaner-tutor interaction model implemented in a Web-based virtual education system. Also, we proposed a new tutor model as a strategy for solving problems found in learner-tutor intraction in Web-based virtual education. Then, based on the results of previous research, we derived the roles of the tutor, and using the roles, modeled components of interaction between the tutor based on existing Web-based virtual education systems to demonstrate the effect. The results showed that the proposed model is efficient for improving the learners' achievements and interactive activities.

• Journal of Korean Ophthalmic Optics Society
• /
• v.9 no.1
• /
• pp.1-9
• /
• 2004

In this study, we presented a research about the development of the phoropter simulation program for education using the 3-D virtual reality in order to study the ophthalmic optics rather efficiently. Especially, by putting the facility such as the refractive error analysis of the eyeball in the cyber tutor inside, we have made the learner to learn and confirm the present situation of refractive correction through refractive power of principal meridians. By organizing the system which induces the active involvement of the learner and showing the result of the execution to the learner intuitively, the implementation of the more efficient education-environment can be possible. Consequently, it is expected that this program will be greatly helpful for the optician training as well as the ophthalmic optics education. As a next phase of study, we will develope the development technique of this simulator more and extend the cyber tutor contents more and make the web service version of this program to be provided through the internet network in order to inspire the learning desire of the learner more and more.

• Proceedings of the Korea Contents Association Conference
• /
• 2007.11a
• /
• pp.719-723
• /
• 2007

It is very active that existing companies of providing e-Learning contents try to differentiate themselves through a business model based on Web 2.0. For instance, Etoos, online education website (www.etoos.com) run by SK Communications has made more space where students can participate in the Web 2.0 era and overhauled its website completely, turning into an open-ended one, which strengthens learning and fun in 2007. This study is to analyze the present state of e-Learning contents with representative e-learning sites for middle and high school students, to find that the development direction for next generation e-Learning lies in developing contents focusing on learners that can get effective feedback and drawing collective intelligence grounded on the essence of Web 2.0, and to suggest 'the project to form virtual private tutor community in e-Learning contents.'

• Journal of Information Processing Systems
• /
• v.16 no.5
• /
• pp.1223-1230
• /
• 2020

Recently, extensive studies have been conducted to apply deep learning to reinforcement learning to solve the state-space problem. If the state-space problem was solved, reinforcement learning would become applicable in various fields. For example, users can utilize dance-tutorial systems to learn how to dance by watching and imitating a virtual instructor. The instructor can perform the optimal dance to the music, to which reinforcement learning is applied. In this study, we propose a method of reinforcement learning in which the action space is dynamically adjusted. Because actions that are not performed or are unlikely to be optimal are not learned, and the state space is not allocated, the learning time can be shortened, and the state space can be reduced. In an experiment, the proposed method shows results similar to those of traditional Q-learning even when the state space of the proposed method is reduced to approximately 0.33% of that of Q-learning. Consequently, the proposed method reduces the cost and time required for learning. Traditional Q-learning requires 6 million state spaces for learning 100,000 times. In contrast, the proposed method requires only 20,000 state spaces. A higher winning rate can be achieved in a shorter period of time by retrieving 20,000 state spaces instead of 6 million.