• The Journal of Korean Association of Computer Education
• /
• v.7 no.6
• /
• pp.117-128
• /
• 2004

Compared with traditional face-to-face instruction, online learning causes learners to experience more severe feeling of isolation and results in higher dropout rate. This is due to the lack of interaction, sense of belonging, membership, interdependency, cooperation among members and social environment that enables persistence in online learning. Therefore, it is very important for grouping e-learning community to lower the dropout rate and eliminate feeling of isolation. In this paper, the research has been done on the inclination test list to be applied for grouping the desirable learning community. And on the basis of this research, the grouping system for e-learning community(GSE) based on intelligent multi agents for an inclination test using homogeneous and heterogeneous items has been developed. GSE system has such properties that construct a personalized user profile by an agent, and then make groupings according to users' inclination. When this system was evaluated, about 88% of learners were satisfied, and they wanted the group not to be disorganized but to be maintained.

• Proceedings of the Korean Information Science Society Conference
• /
• 2004.10a
• /
• pp.85-87
• /
• 2004

유무선 인터넷 기술의 발전은 인터넷을 통한 개인 정보의 효율적인 공유 및 교환을 가능하게 하였다. 최근 이러한 분산 정보의 공유를 위한 네트워킹 기술로 P2P(Peer-to-Peer)가 많은 주목을 받고 있다. 현재 국내외의 많은 대학 및 기관에서 P2P에 관한 연구가 활발히 진행 중 이다. 하지만, 대부분의 P2P 시스템들은 파일공유 위주의 서비스를 제공하며 SETI@HOME을 필두로 한 일부 @HOME 프로젝트들만이 자원 공유 서비스를 제공하고 있다. 그러나 기존의 자원공유 P2P 서비스들은 특정한 목적을 위해 구성됨으로써 자원을 제공하는 일반 사용자는 단순히 자원을 제공할 뿐 그 이상의 역할을 수행할 수가 없다. 이에 본 논문에서는 P2P 시스템에 참여한 모든 사용자가 P2P의 자원 네트워크를 사용할 수 있도록 멀티 에이전트 기반의 자원 공유 P2P 시스템을 제안한다. 일반 사용자는 서비스 생성 프레임워크를 사용하여 자신에게 필요한 테스크 에이전트를 생성할 수 있으며, 스케줄러 및 분배 에이전트, 테스크 에이전트에 의해 수행되어진다. 또한 본 시스템은 group 및 peer의 관리를 위해 특성 학습 에이전트(Specific Learning Agent)의 학습기능을 사용함으로써 P2P가 가지는 불안전한 환경 및 신뢰성 문제를 해결하였다.

• The Transactions of the Korea Information Processing Society
• /
• v.6 no.10
• /
• pp.2618-2629
• /
• 1999

In this paper, we describe the design and implementation of the FDRA(Fault Detection Recovery based on Agent) running on distributed multimedia environment. DOORAE is a good example for distributed multimedia and multimedia distance education system among students and teachers during lecture. It has primitive service agents. Service functions are implemented with objected oriented concept. FDRA is a multi-agent system. It has been environment, intelligent agents interact with each other, either collaboratively or non-collaboratively, to achieve their goals. The main idea is to detect an error by using polling method. This system detects an error by polling periodically the process with relation to session. And, it is to classify the type of error s automatically by using learning rules. The merit of this system is to use the same method to recovery it as it creates a session. FDRA is a system that is able to detect an error, to classify an error type, and to recover automatically a software error based on distributed multimedia environment.

• KIPS Transactions on Software and Data Engineering
• /
• v.12 no.9
• /
• pp.407-418
• /
• 2023

The Multi-Object Goal Visual Navigation(MultiOn) is a visual navigation task in which an agent must visit to multiple object goals in an unknown indoor environment in a given order. Existing models for the MultiOn task suffer from the limitation that they cannot utilize an integrated view of multimodal context because use only a unimodal context map. To overcome this limitation, in this paper, we propose a novel deep neural network-based agent model for MultiOn task. The proposed model, MCFMO, uses a multimodal context map, containing visual appearance features, semantic features of environmental objects, and goal object features. Moreover, the proposed model effectively fuses these three heterogeneous features into a global multimodal context map by using a point-wise convolutional neural network module. Lastly, the proposed model adopts an auxiliary task learning module to predict the observation status, goal direction and the goal distance, which can guide to learn the navigational policy efficiently. Conducting various quantitative and qualitative experiments using the Habitat-Matterport3D simulation environment and scene dataset, we demonstrate the superiority of the proposed model.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.2 no.4
• /
• pp.3-12
• /
• 2009

In recent years, the development of ubiquitous computing environment, according to the time, place, regardless of the environment changes dynamically based on providing a service. In particular, ubiquitous computing environment, education support services in the fields of education according to the principal of each member is required to provide personalized information. Therefore, this paper describes the education support system which provide adaptive information to member of education. The structure of the proposed system consists of mobile agents multi-agent system platform, JADE (Java Agent DEvelopment framework) is based. Also, we describes the design of agents for application services and the interaction model. In this paper, the performance of proposed system to verify availability, classroom teachers, students and parents and administrators as a service application based on the user's role to provide appropriate information system was implemented. Finally, we shows the result of user interface GUIs according to adaptive education services.

• Proceedings of the Korean Information Science Society Conference
• /
• 2006.10b
• /
• pp.393-396
• /
• 2006

강화학습은 환경과 상호작용하는 과정을 통하여 목표를 이루기 위한 전략을 학습하는 방법으로써 에이전트의 학습방법으로 많이 사용한다. 독립적인 에이전트가 아닌 상호 의사소통이 가능한 다중 에이전트 환경에서 에이전트의 학습정보를 서로 검색 및 공유가 가능하다면 환경이 거대하더라도 기존의 강화학습 보다 빠르게 학습이 이루어질 것이다. 하지만 아직 다중 에이전트 환경에서 학습 방법에 대한 연구가 미흡하여 학습정보의 검색과 공유에 대해 다양한 방법들이 요구되고 있다. 본 논문에서는 대상 에이전트 학습 정보와 주변 에이전트들의 학습 정보 사이에 편집거리를 비교하여 유사한 에이전트를 찾고 그 에이전트 정보를 강화학습 사전정보로 사용함으로써 학습속도를 향상시킨 ED+Q-Learning 시스템을 제안한다.

• Annual Conference of KIPS
• /
• 2002.11a
• /
• pp.369-372
• /
• 2002

다양한 특성들을 가지고 있는 멀티에이전트 시스템의 행동학습은 에이전트 설계에 많은 부담을 덜어준다. 특성들로부터 나오는 다양한 행동의 효과적인 학습은 에이전트들이 환경에 대한 자율성과 반응성을 높여준 수 있다. 행동학습은 model-based learning과 같은 교사학습보다는 각 상태를 바로 지각하여 학습하는 강화학습과 같은 비교사 학습이 효과적이다. 본 논문은 로봇축구환경에 에이전트들의 행동을 개선된 강화학습법인 Modular Q-learning을 적용하여 복잡한 상태공간을 효과적으로 나누어 에이전트들의 자율성과 반응성을 높일 수 있는 강화학습구조를 제안한다.

• 제어로봇시스템학회:학술대회논문집
• /
• 2001.10a
• /
• pp.44.1-44
• /
• 2001

This paper discusses the interacting mobile robots for tele-operation system using the world wide web. In multi-agent and web-based teleoperation environment the problem of communication delay must be solved for the efficient and robust control of the system. The standard graphic user interface(GUI)is implemented using Java Programing language. The web browser is used to integrate the virtual environment and the standard GUI(Java applet) in a single user interface. Users can access a dedicated WWWserver and download the user interface. Reinforcement learning is applied to indirect control in order to autonomously operate without the need of human intervention. Java application has been developed to communicate and control multi robots using WWW. The effectiveness of our multi robots system is verified by simulation and experiments ...

• International Journal of Computer Science & Network Security
• /
• v.24 no.10
• /
• pp.71-80
• /
• 2024

The Internet of Things (IoT) is an emerging element that is becoming increasingly indispensable to the Internet and shaping our current understanding of the future of the Internet. IoT continues to extend deeper into the daily lives of people, offering distributed and critical services. In contrast with current Internet, IoT depends on a dynamic architecture where physical objects with embedded sensors will communicate via cloud to send and analyze data [1-3]. Its security troubles will surely impinge all aspects of civilization. Mobile agents are widely used in the context of the IoT and due to the possibility of transmitting their execution status from one device to another in an IoT network, they offer many advantages such as reducing network load, encapsulating protocols, exceeding network latency, etc. Also, cryptographic technologies, like PKI and Blockchain technology, and Artificial Intelligence are growing rapidly allowing the addition of an approved security layer in many areas. Security issues related to mobile agent migration can be resolved with the use of these technologies, thus allowing increased reliability and credibility and ensure information collecting, sharing, and processing in IoT environments, while ensuring maximum autonomy by relying on the AI to allow the agent to choose the most secure and optimal path between the nodes of an IoT environment. This paper aims to present a new model to secure mobile agents in the context of the Internet of Things based on Public Key Infrastructure (PKI), Ethereum Blockchain Technology and Q-learning. The proposed model provides a secure migration of mobile agents to ensure security and protect the IoT application against malevolent nodes that could infiltrate these IoT systems.

• Journal of Advanced Navigation Technology
• /
• v.26 no.4
• /
• pp.226-234
• /
• 2022

With advances in autonomous vehicles and networked control, there is a growing interest in the consensus control of a multi-agents system to control multi-agents with distributed control beyond the control of a single agent. Since consensus control is a distributed control, it is bound to have delay in a practical system. In addition, it is often difficult to have a very accurate mathematical model for a system. Even though a reinforcement learning (RL) method was developed to deal with these issues, it often experiences slow convergence in the presence of large uncertainties. Thus, we propose a slide RL which combines the sliding mode control with RL to be robust to the uncertainties. The structure of a sliding mode control is introduced to the action in RL while an auxiliary sliding variable is included in the state information. Numerical simulation results show that the slide RL provides comparable performance to the model-based consensus control in the presence of unknown time-varying delay and disturbance while outperforming existing state-of-the-art RL-based consensus algorithms.