• Proceedings of the Korea Information Processing Society Conference
• /
• 2024.05a
• /
• pp.666-669
• /
• 2024

연합학습은 원본 데이터를 공유하지 않고 모델을 학습할 수 있는 각광받는 프라이버시를 위한 학습방법론이다. 이를 위해 참여자의 데이터를 수집하는 대신, 데이터를 인공지능 모델 학습의 요소들(가중치, 기울기 등)로 변환한 뒤, 이를 공유한다. 이러한 강점에 더해 기존 연합학습을 개선하는 방법론들이 추가적으로 연구되고 있다. 기존 연합학습은 모델 가중치를 평균내는 것으로 참여자 간에 동일한 모델 구조를 강요하기 때문에, 참여자 별로 자신의 환경에 알맞은 모델 구조를 사용하기 어렵다. 이를 해결하기 위해 지식 증류 기반의 연합학습 방법(Knowledge Distillation-based Federated Learning)으로 서로 다른 모델 구조를 가질 수 있도록(Model Heterogenousity) 하는 방법이 제시되고 있다. 연합학습은 여러 참여자가 연합하기 때문에 일부 악의적인 참여자로 인한 모델 포이즈닝 공격에 취약하다. 수많은 연구들이 기존 가중치를 기반으로한 연합학습에서의 위협을 연구하였지만, 지식 증류 기반의 연합학습에서는 이러한 위협에 대한 조사가 부족하다. 본 연구에서는 최초로 지식 증류 기반의 연합학습에서의 모델 성능 하락 공격에 대한 위협을 실체화하고자 한다. 이를 위해 우리는 GMA(Gaussian-based Model Poisoning Attack)과 SMA(Sign-Flip based Model Poisoning Attack)을 제안한다. 결과적으로 우리가 제안한 공격 방법은 실험에서 최신 학습 기법에 대해 평균적으로 모델 정확도를 83.43%에서 무작위 추론에 가깝게 떨어뜨리는 것으로 공격 성능을 입증하였다. 우리는 지식 증류 기반의 연합학습의 강건성을 평가하기 위해, 새로운 공격 방법을 제안하였고, 이를통해 현재 지식 증류 기반의 연합학습이 악의적인 공격자에 의한 모델 성능 하락 공격에 취약한 것을 보였다. 우리는 방대한 실험을 통해 제안하는 방법의 성능을 입증하고, 결과적으로 강건성을 높이기 위한 많은 방어 연구가 필요함을 시사한다.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.15 no.1
• /
• pp.65-68
• /
• 2005

Privacy preservation is crucial in ubiquitous computing environment in which lots of privacy- sensitive information can be collected and distributed without appropriate control. The anonymity-based approach is a famous one used for privacy preservation communication, which allows users to use pseudonyms instead of real ID so as not to reveal their identities. This approach is effective in that it can hide the identity of users. However, it makes it difficult to provide ID-based services like buddy service, dangerous area alert, P2P communication in the ubiquitous computing. We proposes a system architecture which enables ID-based services in the ubiquitous computing environment employing anonymity - based privacy - preserving approach.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2008.10a
• /
• pp.532-535
• /
• 2008

The requirement of Digital Culture Content(Movie, Music, Animation, Digital TV, Exhibition and etc.) is increasing so variety and quantity of content is also increasing. The Movie what majority of the digital Content is developing of technology and data. In the result, the efficient retrieval service has required and user want to use a recommendation engine and semantic retrieval methods through the recommendation system. Therefore, this paper will suggest analysing trait element of digital content data, building of retrieval technology, analysing and retrieval technology base on KANSEI vocabulary and etc. For the these, we made a extraction technology of trait element based on semantics and KANSEI processing algorithm based on color information.

• Spatial Information Research
• /
• v.7 no.2
• /
• pp.191-207
• /
• 1999

Data and information necessary for solving the spatial decision making problems are imperfect or inaccurate and most are described by natural language. In order to process these arts of information by the computer, the obscure linguistic value need to be described quantitatively to let and computer understand natural language used by humans. For this , the fuzzy set theory and the fuzzy logic are used representative methodology. So this paper describes the construction of the language model by the natural language that user easily can understand and the logical concepts and construction process for building the fuzzy inference mechanism. It makes possible to solve the space related decision making problems intellectually through structuring and inference used by the computer, in case of the evaluation concern or decision making problems are described inaccurate, based on the inaccurate or indistinct data and information.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.17 no.4
• /
• pp.500-505
• /
• 2007

Current humanoid robot technology has a problem of lacking opened methodology about mechanisms of analysis, design, implementation, and integration for robot development but is focused only on manufacture robot and implementation of technology. In this paper, to overcome problems of humanoid robots that have been shown since and for construction of new structure which satisfy the concept of opening, networking, and modularization that is the development direction of future robot, we proposed morphological and neurological model of human arm and design method of humanoid robot arm based on the each joint design and kinematics model.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.15 no.5
• /
• pp.626-631
• /
• 2005

This paper presents the target object search algorithm under dynamic programming (DP) in the Tree-type maze. We organized an experimental environment with the concatenation Y-shape diverged way, small mobile robot, and a target object. By the principle of optimality, the backbone of DP, an agent recognizes that a given whole problem can be solved if the values of the best solution of certain ancillary problem can be determined according to the principle of optimality. In experiment, we used two different control algorithms: a left-handed method and DP. Finally we verified the efficiency of DP in the practical application using a real robot.

• Informatization Policy
• /
• v.22 no.1
• /
• pp.73-95
• /
• 2015

This research has a purpose to verify whether the Internet diffusion through the Official Development Assistance (ODA) can bring about democratic changes in authoritarian political system of developing countries. This research has found out that IT ODA has 'positive(+)' effect on the Internet diffusion in recipient countries. It seems to support the expectation of donor countries that they can lead the political democratization by constructing infrastructure and promoting the use of the Internet. However, as the impact of IT ODA on the Internet diffusion is not considerable as well as the Internet diffusion has no strong influence on the development of democracy in developing countries, and also each of these aspects of countries looks different, it is hard to see that the democratization hypothesis based on the optimistic technological determinism is reasonable. Therefore, this paper argues that IT ODA must be based on not blind optimism that IT transfer will lead to the political democratization in developing countries, but the social structuralism which is based on distinctiveness and uniqueness of countries. and suggests that the discuss of the effect of IT ODA proceed with the constructive way.

• Journal of Intelligence and Information Systems
• /
• v.21 no.1
• /
• pp.15-28
• /
• 2015

As personal devices and pervasive technologies for interacting with networked objects continue to proliferate, there is an unprecedented world of scattered pieces of contextualized information available. However, the explosive growth and variety of information ironically lead users and service providers to make poor decision. In this situation, recommender systems may be a valuable alternative for dealing with these information overload. But they failed to utilize various types of contextual information. In this study, we suggest a methodology for context-aware recommender systems based on the concept of contextual boundary. First, as we suggest contextual boundary-based profiling which reflects contextual data with proper interpretation and structure, we attempt to solve complexity problem in context-aware recommender systems. Second, in neighbor formation with contextual information, our methodology can be expected to solve sparsity and cold-start problem in traditional recommender systems. Finally, we suggest a methodology about context support score-based recommendation generation. Consequently, our methodology can be first step for expanding application of researches on recommender systems. Moreover, as we suggest a flexible model with consideration of new technological development, it will show high performance regardless of their domains. Therefore, we expect that marketers or service providers can easily adopt according to their technical support.

• Journal of Intelligence and Information Systems
• /
• v.29 no.4
• /
• pp.375-393
• /
• 2023

The demand for Less than Container Load (LCL) has been on the rise due to the growing need for various small-scale production items and the expansion of the e-commerce market. Consequently, more companies in the International Freight Forwarder are now handling LCL. Given the variety in cargo sizes and the diverse interests of stakeholders, there's a growing need for a container loading algorithm that optimizes space efficiency. However, due to the nature of the current situation in which a cargo loading plan is established in advance and delivered to the Container Freight Station (CFS), there is a limitation that variables that can be identified at industrial sites cannot be reflected in the loading plan. Therefore, this study proposes a container loading methodology that makes it easy to modify the loading plan at industrial sites. By allowing the characteristics of cargo and the status of the container to be considered, the requirements of the industrial site were reflected, and the three-dimensional space was manipulated into a two-dimensional planar layer to establish a loading plan to reduce time complexity. Through the methodology presented in this study, it is possible to increase the consistency of the quality of the container loading methodology and contribute to the automation of the loading plan.

• Management & Information Systems Review
• /
• v.31 no.2
• /
• pp.199-230
• /
• 2012

This study introduced trust dimension as a mediating variable and analyzed in order to investigate the empirical mechanism on the process of leader's emotional intelligence which is a emotional factor to produce organizational effectiveness. Based on the empirical analysis, first of all, this study found whether the team leader's emotional intelligence gives an influence on the subordinate's trust in leader. Secondly, this study examined the multidimensionality of trust by selecting subordinate's cognition-based trust and affect-based trust on the team leader as mediating variables between team leaders's emotional intelligence and subordinate's affective commitment, organizational citizenship behavior. I found the theoretical implications based on this study; first, from the view on cognition-based and emotion-based trust, the integrated verification of the process of the trust in leader as a mediating role, second, verification of the relationship between leader's emotional intelligence and trust in leader which was not studied before, third, the methodological try to prevent from the common method bias problem through the evaluation of a team leader on a subordinate's organizational citizenship behavior. This study also suggested the ways to apply those theoretical implications to human resource management.