• International Journal of Internet, Broadcasting and Communication
• /
• 제15권3호
• /
• pp.58-65
• /
• 2023

Bankruptcy is a significant risk for start-up companies, but with the help of cutting-edge artificial intelligence technology, we can now predict bankruptcy with detailed explanations. In this paper, we implemented the Category Boosting algorithm following data cleaning and editing using OpenRefine. We further explained our model using the Shapash library, incorporating domain knowledge. By leveraging the 5C's credit domain knowledge, financial analysts in banks or investors can utilize the detailed results provided by our model to enhance their decision-making processes, even without extensive knowledge about AI. This empowers investors to identify potential bankruptcy risks in their business models, enabling them to make necessary improvements or reconsider their ventures before proceeding. As a result, our model serves as a "glass-box" model, allowing end-users to understand which specific financial indicators contribute to the prediction of bankruptcy. This transparency enhances trust and provides valuable insights for decision-makers in mitigating bankruptcy risks.

• 천문학회지
• /
• 제38권3호
• /
• pp.357-363
• /
• 2005

A deep stellar analysis is introduced for the poorly studied open cluster NGC 2509. The Near-IR database of the digital Two Micron All Sky Survey (2MASS) has been used to re-estimate and refine the fundamental parameters of the cluster, i.e. age, reddening, distance, and diameter. As well as, luminosity function, mass function, total mass, relaxation time, and mass segregation of NGC 2509 have been estimated here for the first time..

• 대한원격탐사학회:학술대회논문집
• /
• 대한원격탐사학회 2007년도 Proceedings of ISRS 2007
• /
• pp.234-237
• /
• 2007

Road networks are important geospatial databases for various GIS (Geographic Information System) applications. Road digital maps may contain geometric spatial errors due to human and scanning errors, but manually updating roads information is time consuming. In this paper, we developed a new road features updating methodology using from multispectral high-resolution satellite image and pre-existing vector map. The approach is based on initial seed point generation using line segment matching and a modified double snake algorithm. Firstly, we conducted line segment matching between the road vector data and the edges of image obtained by Canny operator. Then, the translated road data was used to initialize the seed points of the double snake model in order to refine the updating of road features. The double snake algorithm is composed of two open snake models which are evolving jointly to keep a parallel between them. In the proposed algorithm, a new energy term was added which behaved as a constraint. It forced the snake nodes not to be out of potential road pixels in multispectral image. The experiment was accomplished using a QuickBird pan-sharpened multispectral image and 1:5,000 digital road maps of Daejeon. We showed the feasibility of the approach by presenting results in this urban area.

• International Journal of Internet, Broadcasting and Communication
• /
• 제11권2호
• /
• pp.1-10
• /
• 2019

Research classification of software modules was done to validate the approaches proposed for addressing limitations in existing classification approaches. The objective of this study was to replicate the experiments of a recently published research study and re-evaluate its results. The reason to repeat the experiment(s) and re-evaluate the results was to verify the approach to identify the faulty and non-faulty modules applied in the original study for the prioritization of test cases. As a methodology, we conducted this study to re-evaluate the results of the study. The results showed that binary logistic regression analysis remains helpful for researchers for predictions, as it provides an overall prediction of accuracy in percentage. Our study shows a prediction accuracy of 92.9% for the PureMVC Java open source program, while the original study showed an 82% prediction accuracy for the same Java program classes. It is believed by the authors that future research can refine the criteria used to classify classes of web systems written in various programming languages based on the results of this study.

• 유통과학연구
• /
• 제22권5호
• /
• pp.11-26
• /
• 2024

Purpose: The concept of News-finds-me Perception (NFMP) is gaining increasing scholarly interest due to its wide-ranging findings and implications in digital communications and marketing. From the retail marketing and communication approaches, social media is an effective tool to effectively communicate and persuade customers and stakeholders. Nevertheless, a scarcity of systematic review studies that systematically assemble prior research in the field is recognized. Consequently, this research investigated the Scopus database for articles pertaining to NFMP. Research design, data and methodology: The search was conducted on August 24, 2023, retrieving 46 documents. Following a data-cleaning process, 31 documents remained, providing evidence of the subject area's five-year development. The data was refined with OpenRefine and analyzed with VosViewer. Results: An overview of the subject's expansion is presented, which comprises the most cited documents, authors, organizations, journals, and countries. Furthermore, the investigation examines the influential studies that furnished scientists with essential knowledge and identify the current research trend of the research subject. Conclusions: Based on the results, the study proposes theoretical and practical implications, encouraging academics to further integrate the concept with various communication and marketing theories, as well as the retail marketing context, to gain a better understanding of its complex impacts.

• Journal of Information Processing Systems
• /
• 제20권4호
• /
• pp.535-549
• /
• 2024

Facial expressions (FEs) serve as fundamental components for human emotion assessment and human-computer interaction. Traditional convolutional neural networks tend to overlook valuable information during the FE feature extraction, resulting in suboptimal recognition rates. To address this problem, we propose a deep learning framework that incorporates hierarchical feature fusion, contextual data, and an attention mechanism for precise FE recognition. In our approach, we leveraged an enhanced VGGNet16 as the backbone network and introduced an improved group convolutional channel attention (GCCA) module in each block to emphasize the crucial expression features. A partial decoder was added at the end of the backbone network to facilitate the fusion of multilevel features for a comprehensive feature map. A reverse attention mechanism guides the model to refine details layer-by-layer while introducing contextual information and extracting richer expression features. To enhance feature distinguishability, we employed islanding loss in combination with softmax loss, creating a joint loss function. Using two open datasets, our experimental results demonstrated the effectiveness of our framework. Our framework achieved an average accuracy rate of 74.08% on the FER2013 dataset and 98.66% on the CK+ dataset, outperforming advanced methods in both recognition accuracy and stability.

• Wind and Structures
• /
• 제15권1호
• /
• pp.27-41
• /
• 2012

Wind loads on low-rise buildings in general and residential homes in particular can differ significantly depending upon the laboratory in which they were measured. The differences are due in large part to inadequate simulations of the low-frequency content of atmospheric velocity fluctuations in the laboratory and to the small scale of the models used for the measurements. The imperfect spatial coherence of the low frequency velocity fluctuations results in reductions of the overall wind effects with respect to the case of perfectly coherent flows. For large buildings those reductions are significant. However, for buildings with sufficiently small dimensions (e.g., residential homes) the reductions are relatively small. A technique is proposed for simulating the effect of low-frequency flow fluctuations on such buildings more effectively from the point of view of testing accuracy and repeatability than is currently the case. Experimental results are presented that validate the proposed technique. The technique eliminates a major cause of discrepancies among measurements conducted in different laboratories. In addition, the technique allows the use of considerably larger model scales than are possible in conventional testing. This makes it possible to model architectural details, and improves Reynolds number similarity. The technique is applicable to wind tunnels and large scale open jet facilities, and can help to standardize flow simulations for testing residential homes as well as significantly improving testing accuracy and repeatability. The work reported in this paper is a first step in developing the proposed technique. Additional tests are planned to further refine the technique and test the range of its applicability.

• 한국산업보건학회지
• /
• 제33권4호
• /
• pp.419-426
• /
• 2023

Objectives: This study aims to develop an Occupational Safety and Health (OSH) guide for the safe cleaning of contaminated machinery, equipment, and parts used in the electronics manufacturing process. Methods: A literature review, field investigations, and discussions were conducted. An initial draft of an OSH guide was developed and reviewed by experts with significant experience in maintenance work in the electronics manufacturing process in order to refine the guide. Results: Workers involved in cleaning processes with chemicals, solvents, and abrasive blasting can face exposure to a wide range of chemicals, abrasives, and noise. Identifying potential risks associated with each cleaning technique was an essential first step toward enhancing safety measures. The OSH guide comprises approximately eleven to twelve sections spanning 20-25 pages. It includes engineering and administrative protocols systematically organized to address the necessary actions before, during, and after cleaning tasks, depending on the technique. It is recommended that airline respirator masks be used in conjunction with an air purification system to ensure adherence to air quality standard "D" for atmosphere level. The use of an oil-free air compressor is advised, preferably a stationary model that does not rely on fuel sources like diesel. Conclusions: This OSH guide is designed to protect workers involved in maintenance activity in the electronics industry and aligns with global standards, such as those from the International Organization for Standardization (ISO) and Semiconductor Equipment and Material International, ensuring a higher level of safety and compliance.

• 한국컴퓨터정보학회논문지
• /
• 제19권10호
• /
• pp.135-149
• /
• 2014

본 연구는 과학 교과목 학습에 대한 학생들이 가지는 학습적 어려움을 파악하고 기존 연구와 시중에 사용되는 학습 애플리케이션들을 분석하여 U-러닝 시스템을 제안하였다. 사례조사를 통해 기존 대다수 학습 애플리케이션이 가지는 동영상 기반으로 인한 긴 학습시간의 요구와 개방형 학습콘텐츠의 부족으로 인해 학습자들이 자신의 학습 수준을 파악할 수 없다는 점과 양방향 참여가 어렵다는 문제점, 그리고 학습자 수준을 고려하지 않은 교육 콘텐츠 제공과 같은 문제점들을 발견하였다. 이를 보완하기 위해 단기학습 단위의 소규모 학습 콘텐츠, 개방형 학습 시스템, 강화된 계층적인 학습 콘텐츠 등의 설계요소들을 적용하여 새로운 U-러닝 시스템을 설계하였다. 설계된 시스템을 안드로이드 기반으로 구현하여 학습 대상에게 유익한 과학 교육을 제공하였다. 제안한 U-러닝 시스템이 교육 효과가 있음을 보이기 위해 중학생 3학년을 대상으로 설문조사를 실시하였다. 설문조사에서 기존 학습 애플리케이션과 본 논문에서 제안된 U-러닝 시스템을 모두 이용하게 하였다. 분석 결과로서 일방적인 단방향 학습이 아닌 양방향으로 학습에 참여하고, 학습결과를 공유하여 피드백이 가능함으로써 학습효과가 얻어 질 수 있음을 t-검정으로 확인하였다.

• 지능정보연구
• /
• 제21권4호
• /
• pp.53-77
• /
• 2015

국내 모바일 커머스 시장은 현재 소셜커머스가 이용자 수 측면에서 오픈마켓을 압도하고 있는 상황이다. 산업계에서는 모바일 시장에서 소셜커머스의 성장에 대해 빠른 모바일 시장진입, 큐레이션 모델 등을 주요 성공요인으로 제시하고 있지만, 이에 대한 학계의 실증적인 연구 및 분석은 아직 미미한 상황이다. 본 연구에서는 사용자 리뷰를 바탕으로 모바일 소셜커머스와 오픈마켓의 사용자 이용경험을 비교 분석하는 탐험적인 연구를 수행하였다. 먼저 본 연구는 구글 플레이에 등록된 국내 소셜커머스 주요 3개 업체와 오픈마켓 주요 3개 업체의 모바일 앱 리뷰를 수집하였다. 본 연구는 LDA 토픽모델링을 통해 1만여건에 달하는 모바일 소셜커머스와 오픈마켓 사용자 리뷰를 지각된 유용성과 지각된 편리성 토픽으로 분류한 뒤 감정분석과 동시출현단어분석을 수행하였다. 이를 통해 본 연구는 국내 모바일 커머스 상에서 오픈마켓 이용자들에 비해 소셜커머스 이용자들이 서비스와 이용편리성 측면에서 더 긍정적인 경험을 하고 있음을 증명하였다. 소셜커머스는 '배송', '쿠폰', '할인'을 중심으로 서비스 측면에서 이용자들에게 긍정적인 이용경험을 이끌어내고 있는 반면, 오픈마켓의 경우 '로그인 안됨', '상세보기 불편', '멈춤'과 같은 기술적 문제 및 불편으로 인한 이용자 불만이 높았다. 이와 같이 본 연구는 사용자 리뷰를 통해 서비스 이용경험을 효과적으로 비교 분석할 수 있는 탐험적인 실증연구법을 제시하였다. 구체적으로 본 연구는 LDA 토픽모델링과 기술수용모형을 통해 사용자 리뷰를 서비스와 기술 토픽으로 분류하여 효과적으로 분석할 수 있는 새로운 방법을 제시하였다는 점에서 의의가 있다. 또한 본 연구의 결과는 향후 소셜커머스와 오픈마켓의 경쟁 및 벤치마킹 전략에 중요하게 활용될 수 있을 것으로 기대된다.