• Journal of information and communication convergence engineering
• /
• 제16권4호
• /
• pp.213-220
• /
• 2018

With the exponential developments of social network, their fundamental role as a medium to spread information, ideas, and influence has gained importance. It can be expressed by the relationships and interactions within a group of individuals. Therefore, some models and researches from various domains have been in response to the influence maximization problem for the effects of "word of mouth" of new products. For example, in reality, more than two related social groups such as commercial companies and service providers exist within the same market issue. Under such a scenario, they called social adversaries competitively try to occupy their market influence against each other. To address the influence maximization (IM) problem between them, we propose a novel IM problem for social adversarial players (IM-SA) which are exploiting the social network attributes to infer the unknown adversary's network configuration. We sophisticatedly define mathematical closed form to demonstrate that the proposed scheme can have a near-optimal solution for a player.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2013년도 춘계학술발표대회
• /
• pp.478-479
• /
• 2013

Smart home is one of the emerging domains to come up after advances in home appliances and automation technologies. There are many commercial solutions for smart homes yet many of them have yet to truly exploit the potential of private cloud for low level contextual services and ability to migrate to public cloud for more processing and storage. We propose a private cloud gateway for smart home which will have the ability to sense the new devices, ability to detect context of the situation and act in an appropriate way. It will also record the user logs which will be audited for improvement of the overall system.

• 방송공학회논문지
• /
• 제20권5호
• /
• pp.676-686
• /
• 2015

Light field 카메라를 이용하여 영상을 취득한 후 다양한 응용 프로그램으로 확장이 가능한 4차원 light field 영상은 일반적인 2차원 공간영역(spatial domain)과 추가적인 2차원 각영역(angular domain)으로 구성된다. 그러나 이러한 4차원 light field 영상을 유한한 해상도를 가진 2차원 CMOS 센서로 취득하므로 저해상도의 제약이 존재한다. 본 논문에서는 이러한 4차원 light field 영상이 가지는 해상도 제약 조건을 해결하기 위하여, 4차원 light field 영상에 적합한 딕셔너리 학습 기반(dictionary learning-based) 초해상도(superresolution) 알고리즘을 제안한다. 제안하는 알고리즘은 4차원 light field 영상으로부터 추출한 많은 수의 4차원 패치(patch)들을 바탕으로 딕셔너리를 구성 및 훈련하며, 학습된 딕셔너리를 바탕으로 저해상도 입력 영상의 해상도를 향상시키는 과정을 수행한다. 제안하는 알고리즘은 공간영역과 각영역의 해상도를 동시에 각각 2배 향상시킨다. 실험에 사용된 영상은 상용 light field 카메라인 Lytro에서 취득하였고 기존의 알고리즘과의 비교를 통해 제안하는 알고리즘의 우수성을 검증한다.

• Journal of Microbiology and Biotechnology
• /
• 제19권11호
• /
• pp.1306-1318
• /
• 2009

The filamentous ascomycete Sclerotinia sclerotiorum is well known for its ability to produce a large variety of hydrolytic enzymes. Two $\alpha$-amylases ScAmy54 and ScAmy43 predicted to play an important role in starch degradation were showed to produce specific oligosaccharides essentially maltotriose that have a considerable commercial interest. Primary structure of the two enzymes was established by N-terminal sequencing, MALDI-TOF masse spectrometry and cDNA cloning. The two proteins have the same N-terminal catalytic domain and ScAmy43 derived from ScAmy54 by truncation of 96 amino acids at the carboxyl-terminal region. Data of genomic analysis suggested that the two enzymes originated from the same $\alpha$-amylase gene and that truncation of ScAmy54 to ScAmy43 occurred probably during S. sclerotiorum cultivation. The structural gene of Scamy54 consisted of 9 exons and 8 introns, containing a single 1,500-bp open reading frame encoding 499 amino acids including a signal peptide of 21 residues. ScAmy54 exhibited high amino acid homology with other liquefying fungal $\alpha$-amylases essentially in the four conserved regions and in the putative catalytic triad. A 3D structure model of ScAmy54 and ScAmy43 was built using the 3-D structure of 2guy from A. niger as template. ScAmy54 is composed by three domains A, B, and C, including the well-known $(\beta/\alpha)_8$ barrel motif in domain A, have a typical structure of $\alpha$-amylase family, whereas ScAmy43 contained only tow domains A and B is the first fungal $\alpha$-amylase described until now with the smallest catalytic domain.

• 정보처리학회 논문지
• /
• 제13권9호
• /
• pp.421-428
• /
• 2024

최근 GPT-4, LLaMA와 같은 초거대 언어모델을 활용한 서비스가 공개되며 많은 주목을 받고 있다. 이 모델들은 사용자들의 다양한 질문에 유창하게 응답할 수 있지만, 한국어 데이터의 학습량이 부족하여 한국 문화 및 한국어에 대한 잘못된 정보를 제공할 가능성이 있다. 본 논문에서는 한국어 데이터를 학습한 주요 공개 모델 8개를 선정하여, 5개 분야(한국어 이해 및 문화 영역으로 구성)에 대한 평가 데이터셋을 통해 한국 문화 이해 능력을 평가하였다. 그 결과, 상용 모델인 HyperClovaX가 전 분야에서 가장 뛰어난 성능을 보였으며, 공개용 모델 중에서는 Bookworm이 한국어 구사 능력에서 우수한 성과를 보였다. 또한, 한국어 이해 및 문화와 관련된 부문에서는 LDCC-SOLAR 모델이 뛰어난 성능을 확인할 수 있었다.

• 한국산학기술학회논문지
• /
• 제19권1호
• /
• pp.48-55
• /
• 2018

가상현실(VR) 실내자전거의 사용자경험설계(UXD)를 평가하기 위한 방법이 최근에 개발되었으나 이 방법은 실제 제품을 통한 어떠한 실증적인 비교 평가를 받지 못하였다. 본 연구에서는 이 방법의 효과성을 살펴보고 실제 VR자전거에 적용될 때 나타나는 문제점을 발견하고자 한다. 이를 위해 2가지 유형의 VR자전거, 즉 한국형 가상현실 실내자전거 프로토타입(KoVIB)과 상용화되고 있는 외국산 VirZOOM에 각각 적용하여 전반적인 평가를 하였다. 주요 평가기준은 사용성 분야, 감정적 분야, 사용자가치, 현실감 등 4개가 이용되었다. 사용성 분야는 13개, 감정적 분야는 6개, 사용자가치는 8개, 현실감은 16개 하위 요소로 각각 구성된다. 이들 요소들은 프로토타입 평가를 위해 재해석된다. 그 결과 제품 사이의 강약을 쉽게 파악할 수 있었으며, 프로토타입과 상용품 간에 나타날 수 있는 평가 방법의 문제점들도 발견할 수 있었다. 이를 바탕으로 프로토타입의 장점은 홍보로, 약점은 개선사항으로 용이하게 피드백 되며, 또한 문제점을 반영한 수정된 평가방법이 제안된다.

• Journal of Microbiology and Biotechnology
• /
• 제26권12호
• /
• pp.2206-2213
• /
• 2016

Recently, several studies have revealed that commercial microbial identification systems do not accurately identify the uncommon causative species of candidiasis, including Candida famata, Meyerozyma guilliermondii, and C. auris. We investigated the accuracy of species-level identification in a collection of clinical isolates previously identified as C. famata (N = 38), C. lusitaniae (N = 1 2), and M. guilliermondii (N = 5) by the Vitek 2 system. All 55 isolates were re-analyzed by the Phoenix system (Becton Dickinson Diagnostics), two matrix-assisted laser desorption ionization-time of flight mass spectrometry analyzers (a Vitek MS and a Bruker Biotyper), and by sequencing of internal transcribed spacer (ITS) regions or 26S rRNA gene D1/D2 domains. Among 38 isolates previously identified as C. famata by the Vitek 2 system, the majority (27/38 isolates, 71.1%) were identified as C. tropicalis (20 isolates) or C. albicans (7 isolates) by ITS sequencing, and none was identified as C. famata. Among 20 isolates that were identified as C. tropicalis, 17 (85%) were isolated from urine. The two isolates that were identified as C. auris by ITS sequencing originated from ear discharge. The Phoenix system did not accurately identify C. lusitaniae, C. krusei, or C. auris. The correct identification rate for 55 isolates was 92.7% (51/55 isolates) for the Vitek MS and 94.6% (52/55 isolates) for the Bruker Biotyper, as compared with results from ITS sequencing. These results suggest that C. famata is very rare in Korea, and that the possibility of misidentification should be noted when an uncommon Candida species is identified.

• International Journal of Knowledge Content Development & Technology
• /
• 제8권1호
• /
• pp.79-89
• /
• 2018

We collected and analyzed data from e3data.org, which is a global registry of data repository services. We analyzed data profile for three leading Asian economies-Korea, China, and Japan-against the reference data for other participating countries. In particular, we examined how individual countries contribute to the repository, organizational type, versioning and product quality management, and subject tagging. We come to the conclusion that all three Asian countries still fall short in terms of involvement. As for participating institutions, there are 7 from Korea, 64 from China, and 120 from Japan. Among Chinese organizations, 3 are profit, 61 non-profit, and 37 organizations (which yields 1.8%) are involved in repository building. In Japan, there is 1 is commercial and 119 non-profit organizations, of which 57 (3.0%) are involved in repository building. All 7 organizations from Korea are non-profit, and 6 of them (0.3%) are involved in repository building. As regards versioning and product quality management, Korea, China, and Japan are up to par with other countries. Subject analysis reveals that Korea contributes more to geosciences, Japan to physics and geosciences, while China, unlike Korea and Japan, is more active in life sciences. It is hoped that this study will help planning domestic infrastructure for research data repositories with proper consideration for specific research domains and national characteristics.

• 대한조선학회논문집
• /
• 제47권3호
• /
• pp.377-387
• /
• 2010

In this study, the turbulent free surface around KVLCC1 employed in the circular motion test simulation is numerically calculated using a commercial CFD(Computational Fluid Dynamics) code, FLUENT. Also, hydrodynamic forces and yaw moments around a ship model are calculated during the steady turning. Numerical simulations of the turbulent flows with free surface around KVLCC1 have been carried out by use of RANS equation based on calculation of hydrodynamic forces and yaw moments exerted upon the ship hull. Wave elevation is simulated by using the VOF method. VOF method is known as one of the most effective numerical techniques handling two-fluid domains of different density simultaneously. Boundary layer thickness and wake field are changed various yaw velocities of ship model during the steady turning. The calculated hydrodynamic forces are compared with those obtained by model tests.

• 한국소음진동공학회:학술대회논문집
• /
• 한국소음진동공학회 1998년도 춘계학술대회논문집; 용평리조트 타워콘도, 21-22 May 1998
• /
• pp.34-39
• /
• 1998

In order to compute the forces which are transmitted through rubber bushings with a commercial multibody dynamic analysis (MBDA) program, a rubber bushing model is needed. The rubber bushing model of MBDA programs such as DADS or ADAMS is the Voigt model which is simply a parallel spring-viscous damper system, meaning that the damping force of the Voigt model is proportional to the frequency. However, experiments do not necessarily support this proportionality. Alternatively, the viscoelastic characteristics of rubber bushings can be better represented by the complex stiffness. The purpose of this paper is to develop a viscoelastic rubber bushing model for the MBDA programs. Firstly, a methodology is proposed to calculate the complex stiffness of rubber bushings considering static and dynamic load conditions. Secondly, a viscoelastic rubber bushing model developed which uses standard elements provided by DADS. The proposed methods are applied to the rubber bushings of the lower control arms of a rear suspension of a 1994 Ford Taurus model. Then, the forces computed for the rubber bushing model are analyzed and compared with the Voigt model in time and frequency domains.