• The Journal of Korean Association of Computer Education
• /
• v.11 no.1
• /
• pp.97-107
• /
• 2008

The result verification that determines correctness for the work results calculated in each PC is one of the most important issues in PC grid environments. In this literature, voting-based and trust-based schemes have been mainly used to guarantee the correctness of work results. However, these schemes suffer from both waste of resource utilization and high computation delay because they can not effectively cope with dynamic computational environments. To overcome these shortcomings, we introduce the distribution information of PC resources based on credibility and availability into result verification phase. Using this information, we propose a new result verification scheme, which can determine the correctness of work results by each PC resources' credibility and cope with the dynamic changing environments by each PC resources' availability. To demonstrate the efficiency of our result verification scheme, we evaluate the performance of our scheme from the viewpoints of turnaround time and resource utilization, utilizing resource distribution information in the Korea@Home that is a representative PC grid system in domestic. We also compare the performance of our scheme with that of other ones.

• Journal of the Korea Society of Computer and Information
• /
• v.6 no.3
• /
• pp.113-121
• /
• 2001

The employee's are one of the most important parameters determining the success of orga as they have profound influence on the existence, growth and the development of an orga The importance of the workers in the credit unions prompted studies on the human re management. The credit unions are currently facing the impending needs not only to retain t also to help them grew in the organization, as their business prospects are becoming incr shaky. Therefore, this study was conducted to provide suggestion for a future research by estab human relations among the staff members in terms of factors affecting the quality of personal they provide centering on the influence caused by such factors. According1y this study presented the method to improve personal service provided b employees of the credit union by working out problem revealed by previous researches and more developed study results

• Proceedings of the Korean Information Science Society Conference
• /
• 2001.10a
• /
• pp.229-231
• /
• 2001

최근들어 경제활동의 증가로 대부분의 성인들은 몇 장의 신용카드를 소지하고 있을 것이다. 이에 따른 신용카드의 도난 분실 사고는 카드사의 문제가 되고있다. 기존의 탐지 시스템은 도난신고 등의 일반적인 탐지와 갑작스런 사용 액수의 증가를 탐지하여 도난 분실 카드를 판별하였다. 이것은 소액의 부정거래탐지가 어렵다는 단점이 있다. 본 논문에서 제시하는 탐지 시스템은 outlier 기법을 사용하여 training set을 만들고 시간가중치와 거리기반 도표를 이용하여 도난 분실 카드를 탐지한다. 금액, 시간 도표에서 거래요구시간의 차를 계산하여 가중치를 주고 장소, 소비종류 도표에서는 training set에서 얻은 자료인 저녁 8시를 기준으로 소비종류의 배열을 바꾼다. 제안된 시스템은 소액의 부정거래 탐지에도 우수하고 이전의 시스템보다 정확함을 장점으로 한다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2019.05a
• /
• pp.390-393
• /
• 2019

은행과 신용카드 업계에 있어, 고객의 다음 신용 카드 사용처(다음 방문 가맹점)를 예측할 수 있다면 고객의 라이프 스타일을 파악 할 수 있으며, 여러 프로모션과 비즈니스 기회를 포착할 수 있어 매출 증대를 꾀할 수 있다. 우리가 제안하는 모델은 고객이 다음에 방문할 가맹점을 예측/추천하는 것을 목표로 한다. 가맹점 방문과 같이 순차적으로 발생하는 이벤트에는 노이즈가 있을 수 있다. 이 노이즈를 제거하기 위해 두 개의 신경망을 이용한 DoubleDNN을 제안한다. 실험은 BC카드사의 데이터분포를 따르는 인공 생성된 신용카드 사용내역 데이터를 이용하였으며, DoubleDNN은 기존의 다른 추천 모델보다 좋은 성능을 보였다.

• Knowledge Management Research
• /
• v.21 no.3
• /
• pp.229-247
• /
• 2020

In the remarkable growth of P2P financial platform in the field of knowledge management, only companies with big data and machine learning technologies are surviving in fierce competition. The ability to analyze borrowers' credit is most important, and platform companies are also recognizing this capability as the most important business asset, so they are building a credit evaluation system based on artificial intelligence. Nonetheless, online P2P platform providers that offer related services only act as intermediaries to apply for investors and borrowers, and all the risks associated with the investments are attributable to investors. For investors, the only way to verify the safety of investment products depends on the reputation of P2P companies from newspaper and online website. Time series information such as delinquency rate is not enough to evaluate the early stage of Korean P2P makers' credit analysis capability. This study examines the credit analysis procedure of P2P loan platform using artificial intelligence through the case analysis method for well known the top three companies that are focusing on the credit lending market and the kinds of information data to use. Through this, we will improve the understanding of credit analysis techniques through artificial intelligence, and try to examine limitations of credit analysis methods through artificial intelligence.

• Asia-Pacific Journal of Business Venturing and Entrepreneurship
• /
• v.14 no.1
• /
• pp.139-149
• /
• 2019

The purpose of this study is to contribute to the minimization of the social cost due to the insolvency by improving the success rate of the startups by providing useful information to the founders and the start-up support institutions through analysis of non-financial information affecting the failure of the startups. This study is aimed at entrepreneurs. The entrepreneurs that are defined by the credit guarantee institutions generally refer to entrepreneurs within 5 years of establishment. The data used in the study are sampled from the companies that were supported by the start-up guarantee from January 2014 to December 2013 as the end of December 2017. The total number of sampled firms is 2,826, 2,267 companies (80.2%), and 559 non-performing companies (19.8%). The non-financial information of the entrepreneur was divided into the entrepreneur characteristics information, the entrepreneur characteristics information, the entrepreneur asset information and the entrepreneur 's credit information, and cross-tabulations and logistic regression analysis were conducted. As a result of cross-tabulations, univariate analysis showed that personal credit rating, presence in the industry, presence of residential housing, presence of employees, and presence of financial statements were selected as significant variables. As a result of the logistic regression analysis, three variables such as personal credit rating, occupation in the industry, and presence of residential house were found to be important factors affecting the failure of founding companies. This result shows the importance of entrepreneur 's personal credibility and experience and entrepreneur' s assets in business management. The start-up support institutions should reflect these results in the entrepreneur 's credit evaluation system, and the entrepreneurs need training on the importance of the personal credit and the management plan in the entrepreneurial education. The results of this analysis will contribute to the minimization of the incapacity of startups by providing useful non-financial information to founders and start-up support organizations.

• Journal of the Korea Convergence Society
• /
• v.10 no.4
• /
• pp.175-182
• /
• 2019

The purpose of this study is to test whether credit rating agencies highly evaluate the credit ratings of corporate bonds based upon management participation and ownership share by family governance in ownership structure forms. The samples of this study for empirical analysis were 1,449 non-financial companies listed on Korean Exchange from 2011 to 2016, over whose firm/year data this study conducted regression analysis. The results of empirical analysis in this study are as follows. First, family businesses had positive effects on the evaluation of corporate credit ratings. Second, if the ownership share of family businesses was higher, corporate credit ratings were higher. This result means that high ownership share in family businesses has very positive effects on the credit ratings of related businesses. It is meaningful that this study tested the effect that family businesses can alleviate agency problems and reduce information asymmetry. Furthermore, it is also academically meaningful that this study can contribute to future studies on the role of ownership structure.

• Proceedings of the KAIS Fall Conference
• /
• 2009.12a
• /
• pp.395-397
• /
• 2009

본 연구는 벤처 또는 기술 기반 중소기업을 대상으로 기업의 신용등급에 영향을 미치는 요인을 분석하고 그 시사점을 제시하고자 하였다. 요인분석 결과에 따르면 벤처 기술 지향형 중소기업이 신용등급의 수준을 높이기 위해서는 기존의 재무적 성과만이 아니라 경영일반, 마케팅 및 인적자원관리, 연구개발, 생산 및 품질관리 등의 다양한 경영활동을 효과적으로 관리해 나갈 필요가 있음을 시사한다. 본 연구의 결과는 향후 벤처 기술 기반 중소기업의 경영성과 및 정부지원 사업의 정책적 효과를 평가함에 있어서 유용한 정보가 될 것으로 기대한다.

• Asia-Pacific Journal of Business Venturing and Entrepreneurship
• /
• v.15 no.2
• /
• pp.193-208
• /
• 2020

In this study, we investigate the effects of technology innovation investments and employment on credit ratings of very small start-up businesses using the data period of 2009 till 2015 test two hypotheses: asymmetric information hypothesis or competence hypothesis. We use financial and non-financial data of 51,903 observations of 12,028 small businesses from a database of a commercial bank and fixed effects panel models and two-stage instrumental variable models. We find that in the short-run small size startups show lower credit ratings than non-startups, and that both technology innovation activities and employment capability improve their credit ratings. In the long-run, technology innovation investments do not improve their credit ratings of later years while employment capability improve their credit ratings of the subsequent year. In addition, the age of startups improves their credit ratings of the current year and until the subsequent two years while employee productivity, fixed ratio and ROA positively affect their credit ratings for up to three years. However, short-term and overall debt ratios, cost of borrowings and firm-size negatively affect their credit ratings for up to three years. The results of the study on credit ratings suggest that credit rating agencies seem to consider both technology innovation activities and employment capability in the credit ratings of small start-ups as 'competence factors' rather than 'asymmetric information factors' with inefficiency and cost burdens. The results also suggest that we must find ways to reflect properly the severe asymmetric information of the early-stage start-ups, and technology innovation activities and employment capability in the credit rating formula.

• The Journal of Information Systems
• /
• v.15 no.3
• /
• pp.175-186
• /
• 2006

최근 정보기술(IT; Information Technology) 및 네트워크 기술의 발달은 기업사회의 의사결정 패턴에 큰 변화를 주고 있다. 특히 글로벌 정치경제 환경이 급변함에 따라 기업들의 의사결정은 보다 빠른 피드백(feedback loop)을 요구하고 있어 과거의 정확성을 중심의 패턴에 변화된 정보의 시기적 절성(timely information)이 크게 강조되고 있다. 본 논문에서는 이러한 첨단기술사회에서 빠르게 의견수렴을 할 수 있는 기술적인 프레임워크를 구축하였다. 본 시스템은 현대사회의 주요한 경제 및 재무의사결정 구조(infrastructure)인 신용평가(credit rating)제도를 웹 기반 시스템으로 구현함으로서 정보의 시기적절성과 현재성을 높이는 의사결정지원시스템을 시현하였다.