A5100452179- Educational Systems and Policies
- Technology and Data Analysis
- Internet of Things and Social Network Interactions
- Embedded Systems Design Techniques
- Educational Research and Pedagogy
- Advancements in PLL and VCO Technologies
- Innovation in Digital Healthcare Systems
- Advanced Software Engineering Methodologies
- Software Testing and Debugging Techniques
- Speech and Audio Processing
- Software Reliability and Analysis Research
- Video Coding and Compression Technologies
- Network Time Synchronization Technologies
- Service-Oriented Architecture and Web Services
- Speech Recognition and Synthesis
- Image and Video Quality Assessment
- Recommender Systems and Techniques
- Education and Learning Interventions
- Software System Performance and Reliability
- Advanced Graph Neural Networks
- Natural Language Processing Techniques
- Energy and Environmental Systems
- Cryptography and Residue Arithmetic
- Bayesian Methods and Mixture Models
- Energy Efficient Wireless Sensor Networks
Sejong University
2015-2024
University of Colorado System
2024
University of Colorado Boulder
2024
Korea Institute of Science and Technology
2022-2023
Korea University
2010-2023
Electronics and Telecommunications Research Institute
2022-2023
Hanyang University
2011-2022
Seoul St. Mary's Hospital
2022
Daejeon St. Mary's Hospital
2022
Catholic University of Korea
2012-2022
This paper presents an overview of yield, reliability, burn-in, cost factors, and fault coverage as practiced in the semiconductor manufacturing industry. Reliability yield modeling can be used a foundation for developing effective stress which turn warranty high-quality products. Yield models are described their advantages disadvantages discussed. Both reliability relationships relation between thoroughly analyzed regard to importance
Text-to-video generation marks a significant frontier in the rapidly evolving domain of generative AI, integrating advancements text-to-image synthesis, video captioning, and text-guided editing. This survey critically examines progression text-to-video technologies, focusing on shift from traditional models to cutting-edge Sora model, highlighting developments scalability generalizability. Distinguishing our analysis prior works, we offer an in-depth exploration technological frameworks...
This study evaluates the quality of current schedule a professional soccer league, K League 1 (Korea Professional Soccer League), which is regarded as top Asian league and has unique characteristics compared to European leagues, such odd numbers games between some teams final rounds based on previous performance. It assesses total travel distance play all scheduled games, match fairness, attendance fairness. Match fairness implies that each team should have opportunity against other under...
Conversational Recommender Systems (CRS) aim to provide tailored recommendation responses via a chat interface, including both the user's preferred item and its accompanying explanation. However, due generative nature, CRS are prone responding with factually incorrect explanations (i.e., hallucinations). To solve this problem, we propose incorporating passage retrieval module into objective of enhancing factuality informativeness system responses. Specifically, outline essential directions...
The elevated activity of leucine-rich repeat kinase 2 (LRRK2) is implicated in the pathogenesis Parkinson's disease (PD). quest for effective LRRK2 inhibitors has been impeded by formidable challenge crossing blood-brain barrier (BBB). We leveraged structure-based de novo design and developed robust three-dimensional quantitative structure-activity relationship (3D-QSAR) models to predict BBB permeability, enhancing likelihood inhibitor's brain accessibility. Our strategy involved synthesis...
Software architecture is a core asset for any organization that develops software-intensive systems. Unsuitable can precipitate disaster because the determines structure of project. To prevent this, software must be evaluated. The current evaluation methods, however, focus on single product architecture, and not line architectures they hardly consider characteristics lines, such as variation points. This paper describes extension scenario-based analysis technique architecture-called EATAM,...
Online news providers such as Google News, Bing and NAVER News collect a large number of articles from variety presses distribute these to users via their portals. Dynamic nature domain causes the problem information overload that makes it difficult for user find her preferable articles. Motivated by this situation, Corp., largest portal company in South Korea, identified four design considerations (DCs) recommendation reflect unique characteristics domain. In paper, we introduce large-scale...
Background . The aim of this study was to compare the effect high-dose oral rabeprazole versus IV PPI on rebleeding after endoscopic treatment bleeding peptic ulcers. Methods This a two-center, prospective, randomized, controlled trial. Patients with high-risk ulcer had hemostasis and were randomly assigned group (20 mg twice daily for 72 hours) or omeprazole (80 as bolus injection followed by continuous infusion at 8 mg/h hours). Results stopped because slow enrollment (total<mml:math...
Fine-tuning pre-trained large language models (LLMs) with limited hardware presents challenges due to GPU memory constraints. Various distributed fine-tuning methods have been proposed alleviate constraints on GPU. However, determining the most effective method for achieving rapid while preventing out-of-memory issues in a given environment remains unclear. To address this challenge, we introduce LLMem, solution that estimates consumption when applying across multiple GPUs and identifies...
The aim of this study was to investigate the factors associated with decrease in speech discrimination ability seen patients presbycusis and whether aging has a significant effect on observed decline.We retrospectively analyzed medical records aged >40 years who had undergone pure-tone audiometry, auditory brainstem response for hearing loss at our hospital from January 2019 June 2021, investigated that correlated score.We enrolled 103 120 ears, mean age 65.8±11.9 years. average patients'...
This paper presents an efficient method in deciding macro-block mode and selecting prediction for intra H.264/AVC high profile. supports nine modes luminance 4x4 8x8 blocks. The predictors include 8 directional DC mode, which is non-directional mode. For 16x16 blocks, the following 4 are used: Vertical, Horizontal, Plane To select best rate-distortion costs of all possible calculated exhaustively. Thus, process has computational complexity. Especially, profile, it extremely large since...
Software architectural design is becoming increasingly important in software engineering, as being manifested through various recent developments the field such component-based engineering paradigm and distributed collaborative computing paradigm. Abstraction a mechanism key concept underpinning architecture, namely hiding immense amount of details. Despite its long-recognized benefits, however abstraction can also pose difficulties with understanding analysis architecture since one result...
최근 클라우드 컴퓨팅은 인터넷 접속을 통해 언제어디서든 사용할 수 있는 높은 이용편리성과 동시에 스마트폰, 넷북, PDA 등과 같은 각종 정보통신 기기로 데이터를 손쉽게 공유할 사용환경을 제공하기 때문에 산업적 파급효과가 커 디지털혁명을 주도할 서비스로 주목받고 있다. 이와 컴퓨팅 기반의 협업 시스템을 비즈니스 실무부서 간의 업무 통합이 점차적으로 이루어지면, 관련 부서 간 공유하게 되는 데이터가 더욱 많아지기 실무자가 필요한 보다 쉽게 찾아 방법이 필요하다. 기존 연구에서는 군집화를 탐색과정을 단순화했지만, 본 논문에서는 간에 자주 발생하는 데이터 중복을 제거하고 시스템 성능을 향상시키기 위해 해쉬함수를 사용하고, 변경된 데이터에 대한 정보가 동적으로 반영되어 실무자에게 적합한 분류될 있도록 마이닝 기법 중 베이지안 네트워크를 사용한 제안하였다. 시스템은 방법과 비교하여 탐색기능이 향상된 결과를 나타내었을 뿐만 아니라, CPU, Network Bandwidth 사용량 등의...
In a news recommender system, user tends to click on article if she is interested in its topic understood by looking at title. Such behavior possible since, when viewing the title, humans naturally think of contextual meaning each title word leveraging their own background knowledge. Motivated this, we propose novel personalized recommendation framework CAST (Context-aware Attention network with Selection module for Title representation), which capable enriching words body text that fully...
Fine-tuning pre-trained large language models (LLMs) with limited hardware presents challenges due to GPU memory constraints. Various distributed fine-tuning methods have been proposed alleviate constraints on GPU. However, determining the most effective method for achieving rapid while preventing out-of-memory issues in a given environment remains unclear. To address this challenge, we introduce LLMem, solution that estimates consumption when applying across multiple GPUs and identifies...
이 글에서는 "세벌식" 한글 자판의 지지자들이 개인용 컴퓨터(PC) 시대로 접어들면서 세벌식 자판을 옹호하는 논리를 어떻게 시대에 맞게 다듬어 나갔는지 보일 것이다. 기계식 타자기를 발명했던 공병우(1907~1995)는 시장을 선점하고 많은 지지자들을 확보했으나, 1969년 그의 자판이 정부 표준으로 인정받지 못하면서 어려움을 겪었다. 그리고 1983년 두벌식 새롭게 컴퓨터용 표준자판으로 제정되면서 자판은 잊힐 위기에 처했다. 하지만 공병우와 지지자들은 컴퓨터 맞는 새로운 옹호 담론을 개발함으로써 위기를 헤쳐 나가고 사용자를 확보하고자 했다. 특히 자판으로 입력할 때 글이 진행하면서 자음의 위치가 종성과 초성을 오가는 현상을 심각한 문제로 규정하고 "도깨비불 현상"이라는 이름을 붙여서 공세의 초점으로 삼았다. 두벌식을 비판하고 세벌식을 이들의 담론은 1990년대 초반 유행하던 PC통신 등을 무대로 삼아, 기성세대에 비판적이고 대항문화에 우호적이었던 대학생을 비롯한 젊은...