• Journal of Korea Multimedia Society
• /
• v.11 no.2
• /
• pp.175-181
• /
• 2008

A 2 dimensional quality function for quantifying the quality of images observed at viewing regions in contact-type multiview 3 dimensional imaging systems is devised and its effectiveness is verified experimentally. The function allows calculating the locations of viewing regions for patched images of any number of different view images. The images observed at the calculated regions reveal that different parts of the display panel of the system project different view images and match with the function's prediction.

• Annual Conference on Human and Language Technology
• /
• 2009.10a
• /
• pp.225-230
• /
• 2009

본 논문에서는 중국어의 이메일 특성을 이용한 중한 대화체 자동번역 방법에 대하여 기술한다. 본 논문에서는 중국어와 한국어와 같이 언어 간의 어순이 다르고 이메일과 같이 특정한 도메인의 언어적 자원도 제한적인 특성을 고려하여 중국어 이메일 특성을 이용한 규칙 기반의 번역 방법을 시도하였다. 이를 위해, 본 논문에서는 중국어의 굳어진 표현이 많고, 한글자 단어 많으며, 입력 오류 많고, 청유 및 경어가 많은 이메일 특성 분석을 통해 그에 대응되는 처리 방법을 제안하였다. 그리고, 그 방법의 타당성을 증명하기 위해 규칙기반의 중한 뉴스 자동번역 시스템과 비교 실험을 하였으며, 규칙기반과 통계적 방법의 타당성 실험을 위해 Gmail과도 비교 실험을 하였다. 두 가지 비교 실험 결과, 본 논문에서 접근한 방법이 모두 우수하였으며, 그 타당성을 증명하였다.

• Geotechnical Engineering
• /
• v.13 no.6
• /
• pp.147-164
• /
• 1997

This study has the assumption that scale effects in rock mass properties are atrributed to the discontinuous and inhomogeneous nature of rock masses. In order to escape the general equivalent material approach applied to the concept of representative volume element, this study presents the new method considering irregular i oink geometry and arbitrary numbers of i oink and arbitrary joint orientations. Based on the theoretical approach, this theory is applied to a real engineering project. Showing the property variations with size of rock mass element, various numerical experiments about scale effect are conducted. Particularly, to prove the adequacy of the verification process in scale effect with nomerical method, and to investigate the detailed source of scale effect, 4 models with increas ins number of joints are tested. On the basis of the experimental results, the test results of scale effects in 3-D rock mass are presented. From these experiments the effects of the mechanical properties of rock joints on the scale effects in rock mass strength and elastic constants are discussed. To verify the mechanism of scale effects in jointed rock mass, two models with different j oink geometries are studied.

• Proceedings of the Korean Information Science Society Conference
• /
• 1998.10c
• /
• pp.9-11
• /
• 1998

Ripple Down Rules(RDR)이론은 지식베이스시스템을 지식공학구축기술 또는 지식공학자의 도움 없이 특수분야전문가에 의해 효율적으로 유지보수, 구축되어진다. 특히 시스템의 운용 중 지식베이스의 수정을 효율적으로 처리할 수 있다. 본 논문은 단일결론을 생성하는 RDR이론의 확장인 복수(複數)결론(multiple classification)을 유도하는 MCRDR이론에 대하여 설명한다. MCRDR은 복잡한 복수결론을 허락하면서 RDR이론의 최대 장점인 지식베이스의 간편한 유지보수 기증을 유지한다. MCRDR의 KA과정, 기초케이스 문제해결방법, 그리고 복수결론 추론문제에 대하여 논할 것이다. MCRDR시스템의 우수성을 모의전문가를 이용한 시스템 수축과 실험으로 증명해 보일 것이다. 이 실험을 통하여 복수결론을 지원하는 MCRDR이론이 단일결론을 지원하는 RDR이론을 통하여 효율적으로 증명하고, 또한 기존의 기계학습방법과의 차이점도 보여줄 것이다.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.29 no.6
• /
• pp.1285-1303
• /
• 2019

In the current credit scoring system, the credit bureau gathers credit information from financial institutions and calculates a credit score based on it. However, because all sensitive credit information is stored in one central authority, there are possibilities of privacy violations and successful external attacks can breach large amounts of personal information. To handle this problem, we propose privacy-preserving credit scoring in which a user gathers credit information from financial institutions, calculates a credit score and proves that the score is calculated correctly using a zero-knowledge proof and a blockchain. In addition, we propose a zero-knowledge proof scheme that can efficiently prove committed inputs to check whether the inputs of a zero-knowledge proof are actually provided by financial institutions with a blockchain. This scheme provides perfect zero-knowledge unlike Agrawal et al.'s scheme, short CRSs and proofs, and fast proof and verification. We confirmed that the proposed credit scoring can be used in the real world by implementing it and experimenting with a credit score algorithm which is similar to that of the real world.

• The KIPS Transactions:PartA
• /
• v.11A no.2
• /
• pp.203-206
• /
• 2004

Wegmann's method has been known as the most efficient one for the Theodorsen equation that is needed to solve conformal mapping. It was researched in the earlier studies (1). However divergence was revealed in some difficult problems by numerical experiment using Wegmann's method. We analyzed the cause of divergence and proposed an improved method by applying a low frequency pass filter to Wegmann's method. Numerical experiments using the improved method showed convergence for all divergent problems using the Wegmann's method. In this paper, we prove theroretically the cause of convergence in the Numerical experiment using the improved method by applying a low frequency pass filter to Wegmann's method. We make use of Fourier transforms in this theoretical proof of convergence.

• Proceedings of the Korean Information Science Society Conference
• /
• 2001.04b
• /
• pp.364-366
• /
• 2001

문서 범주화는 전자 정보환경에서 매우 유용한 정보처리 도구로서, 다양한 문서 범주화 기법 및 성능향상을 위한 연구들이 지속적으로 이루어지고 있다. 그러나, 대부분의 연구들은 문서 범주화의 대상이 되는 단어 자질 공간의 차원축소 문제에만 집중되었을 뿐, 학습단계에 큰 영향을 미치는 다범주 단어 자질의 범주 모호성은 고려하지 않았다. 본 연구에서는, 다범주 자질의 범주 모호성을 해소함으로써 문서 범주화의 성능향상을 유도하는 범주 모호성 해소 가중치 W를 제시하고 이를 실험을 통해 증명하였다. 실험에서는 Naive Bayes 분류기와 가중치 W를 적용한 Naive Bayes-W 분류기를 직접 구축하여 문서 범주화의 성능향상 여부를 비교하는데 사용하였다. 도출된 실험결과를 통해, 가중치 W는 현재의 분류기가 가지고 있는 자질 표현의 범주 모호성이라는 단점을 보완하고 분류기의 성능향상을 유도함으로써 정보검색시스템의 검색효율을 높이는 데 활용될 수 있음일 증명되었다.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.33 no.11A
• /
• pp.1053-1062
• /
• 2008

This paper provides a general and compact expression for the probability density function (pdf) and the moment-generating function (MGF) of the maximal ratio combiner output over Rayleigh fading channels. It is then used to derive closed form expression bit error rate (BER) for M-PSK, M-PAM and M-QAM, respectively. A variety of simulations is performed and shows that they match exactly with analytic ones.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2003.10a
• /
• pp.1015-1020
• /
• 2003

Among the techniques to protect private information by adopting biometrics, speaker verification is expected to be widely used due to advantages in convenient usage and implementation cost. Speaker verification should achieve a high degree of the reliability in the verification score, the flexibility in speech text usage, and the efficiency in verification system complexity. Continuants have excellent speaker-discriminant power and the modest number of phonemes in the category, and multilayer perceptrons (MLPs) have superior recognition ability and fast operation speed. In consequence, the two provide viable ways for speaker verification system to obtain the above properties. This paper implements a system to which continuants and MLPs are applied, and evaluates the system using a Korean speech database. The results of the experiment prove that continuants and MLPs enable the system to acquire the three properties.

• The Journal of the Korea Contents Association
• /
• v.19 no.9
• /
• pp.637-645
• /
• 2019

The RFID group proof is an extension of the yoking proof proving that multiple tags are scanned by a reader simultaneously. Existing group proof schemes provide only delayed tag loss detection which detects loss of tag response in a verification phase. However, delayed tag loss detection is not suitable for real-time applications where tag loss must be detected immediately. In this study, I propose a tag response loss detection scheme which detects loss of tag response in the proof generation process quickly. In the proposed scheme, the tag responds with the sequence number assigned to the tag group, and the reader detects the loss of the tag response through the sequence number. Through an experiment for indistinguishability, I show that the sequence number is secure against an analyzing message attack to distinguish between specific tags and tag groups. In terms of efficiency, the proposed scheme requires fewer transmissions and database operations than existing techniques to determine which tags response is lost.