• 정보처리학회논문지:컴퓨터 및 통신 시스템
• /
• 제11권8호
• /
• pp.269-280
• /
• 2022

블록체인 기반 탈중앙화 금융 시장의 등장과 함께 분산 금융 어플리케이션의 사용자와 자본 유동성이 크게 증가하고 있다. 자동화 마켓 메이커는 탈중앙화 거래 플랫폼의 유동성을 바탕으로 자동으로 거래 단가를 계산하는 프로토콜로, 거래 플랫폼의 유동성 풀을 활용하여 구매자와 판매자가 동 시간대에 존재하지 않아도 거래의 성사를 가능하게 하여 현재 탈중앙화 거래 시장에서 가장 일반적으로 사용되고 있다. 그러나 자동화 마켓 메이커는 거래의 비용 효율이 거래에 사용되는 특정 유동성 풀의 유동성 규모에 의존하여, 유동성 풀의 규모가 작고 거래의 규모가 클수록 거래의 비용 효율이 크게 감소한다는 단점이 있다. 이를 해결하기 위한 방법으로 상대적으로 유동성의 규모가 큰 유동성 풀로 거래 경로를 우회하여 비용 효율을 개선하는 거래 경로 라우팅 알고리즘이 채택되고 있지만, 단일 거래 경로만을 사용하기 때문에 여전히 개선의 여지가 존재한다. 본 논문에서는 거래 경로를 우회하는 것에 더하여 거래 규모를 분산하여 다수의 거래 경로를 동시에 탐색하는 다중 경로 라우팅 알고리즘을 제안하고, 해당 알고리즘을 사용하여 자동화 마켓 메이커 기반의 거래 환경에서 기존의 거래 경로 라우팅 알고리즘에 비해 거래의 비용 효율을 개선한다.

• 산업공학
• /
• 제18권1호
• /
• pp.44-51
• /
• 2005

The rapid expansion of the Internet-related technology has changed the current commercial transaction process. In the physical commercial transaction, most deals are accomplished through the negotiation process except a fixed price system. Thus, the negotiation is essential to conclude the business transactions. Especially, under the e-commerce environment, an automated negotiation system is necessary to respond quickly and flexibly to the diverse environmental changes and also to perform negotiations consistently and effectively. To this end, we develop an automated negotiation system using multi-agent technology. This new system includes such functions as creating negotiation proposals automatically, evaluating the counterpart’ proposals, and preparing counter offers.

• 한국IT서비스학회지
• /
• 제18권2호
• /
• pp.143-159
• /
• 2019

Machine learning (ML) is a method of fitting given data to a mathematical model to derive insights or to predict. In the age of big data, where the amount of available data increases exponentially due to the development of information technology and smart devices, ML shows high prediction performance due to pattern detection without bias. The feature engineering that generates the features that can explain the problem to be solved in the ML process has a great influence on the performance and its importance is continuously emphasized. Despite this importance, however, it is still considered a difficult task as it requires a thorough understanding of the domain characteristics as well as an understanding of source data and the iterative procedure. Therefore, we propose methods to apply deep learning for solving the complexity and difficulty of feature extraction and improving the performance of ML model. Unlike other techniques, the most common reason for the superior performance of deep learning techniques in complex unstructured data processing is that it is possible to extract features from the source data itself. In order to apply these advantages to the business problems, we propose deep learning based methods that can automatically extract features from transaction data or directly predict and classify target variables. In particular, we applied techniques that show high performance in existing text processing based on the structural similarity between transaction data and text data. And we also verified the suitability of each method according to the characteristics of transaction data. Through our study, it is possible not only to search for the possibility of automated feature extraction but also to obtain a benchmark model that shows a certain level of performance before performing the feature extraction task by a human. In addition, it is expected that it will be able to provide guidelines for choosing a suitable deep learning model based on the business problem and the data characteristics.

• 한국IT서비스학회지
• /
• 제7권2호
• /
• pp.1-12
• /
• 2008

The development of IT and explosively growing number of Internet users are rapidly spreading and developing e-commerce, while creating diverse on-line transaction methods such as a negotiation, a reverse auction, and a bid. Among these transaction methods, the transactions by means of a negotiation are being made for goods that have no posted price. In particular, the transactions by means of a negotiation are expected to be widely used in the B2B. In order to determine their transportation costs, shippers usually make negotiations with many transporters and logistics companies. And before long, these negotiations are expected to be made on an on-line automated negotiation system. Because of this, this study has tried to develop an automated negotiation methodology that is absolutely necessary for an on-line automated negotiation. This study has estimated and selected the evaluation functions for multi-lateral negotiators' proposals, thus developing an automated negotiation methodology. As a result of this study, a new direction for an automated negotiation has been suggested. Also we expect that this study will be widely used in the automated negotiation of diverse fields.

• 한국통신학회논문지
• /
• 제31권4B호
• /
• pp.285-290
• /
• 2006

엔터프라이즈 애플리케이션 개발을 위한 분산 아키텍처를 개발할 때는 여러 고려 사항 중에서 계층(tier)의 응집력(cohesion)을 높이고 계층간 연결 결합력(coupling)을 낮추기 위해 페이지 작성자와 소프트웨어 개발자의 역할을 명확히 분리하는 것과 비즈니스 로직의 단위가 되는 엔터티(entity)를 정의하고 데이터베이스 연결과 트랜잭션 처리에 엔터티의 사용 및 역할에 대해 정의하는 것이 우선적으로 필요하다. 이 논문에서는 DONSL(Data Server of Non SQL query) 아키텍처를 제시하여 이러한 문제점을 해결하고자 한다. 이 아키텍처는 웹 티어 오브젝트 모델링 방법을 사용하며 계층(tier)간의 결합도를 낮추고, 데이터베이스 연결에 반드시 사용되는 DAO(Data Access Object)와 엔터티를 효과적으로 분리하여 이러한 문제점을 해결 한다. 핵심 내용으로 DAO에서 엔터티 객체를 제거하는 방안을 통해 DAO 개발을 용이하게 하는 방법과 SQL 질의 자동 생성을 통해 트랜잭션 처리 자동화 방법, 그리고 트랜잭션 처리시 AET(Automated Executed Transaction)와 MET(Manual Executed Transaction)를 효율적으로 운용하는 방법에 대해 제시하고 시스템을 구현하였다.

• 통상정보연구
• /
• 제10권1호
• /
• pp.139-154
• /
• 2008

This research want to discuss some issues about payment system of America. This research focuses on comparing ACH with Fedwire. Firstly Fedwire is the payment system for only large-dollar transfer(Large-Value Transfer System; LVTS), while ACH(Automated Clearing House)payment mainly applies a consumer account. Secondly ACH transaction is either credit or debit transfer, while Fedwire is only credit transfer. Thirdly the communication system of Fedwire has twoway method, telephonelike communication network, but ACH is oneway store and forward electronic mail. At present LVTS divides into RTGS(Real-Time Gross Settlement) and DNS(Deferred Net Settlement). However Fedwire uses either RTGS(instantaneous for every transaction) or DNS(net net transaction) but ACH uses only DNS(net net transaction). Fourthly ACH is substantially cheaper than the Fedwire payment. lastly security is important in both Fedwire and ACH, it is even more crucial in Fedwire In addition, for each transaction, Fedwire may entail individual instantaneous confirmation to the originator and notification to the receiver, whereas ACH do not.

• 제어로봇시스템학회논문지
• /
• 제10권11호
• /
• pp.1058-1064
• /
• 2004

This paper presents a distributed precedence queue mechanism to resolve unexpected transmission delay of a lower priority transaction in a CAN based system, which keeps a fixed priority in data transaction. The mechanism is implemented in the upper sub-layer of the data link layer(DLL), which is fully compatible with the original medium access control layer protocol of CAN. Thus the mechanism can be implemented dynamically while the data transactions are going on without any hardware modification. The CAN protocol was originally developed to be used in the automotive industry, and it was recently applied for a broader class of automated factories. Even though CAN is able to satisfy most of real-time requirements found in automated environments, it is not to enforce either a fair subdivision of the network bandwidth among the stations or a satisfactory distribution of the access delays in message transmission. The proposed solution provides a superset of the CAN logical link layer control, which can coexist with the older CAN applications. Through the real experiments, effectiveness of the proposed mechanism is verified.

• 산학경영연구
• /
• 제13권
• /
• pp.207-221
• /
• 2000

신기술에 의한 체제정비가 요구되는 시점에서 1999년 채택된 미국 통일전자거래법은 우리나라에 주는 시사점이 크다. 통일전자거래법은 개인간의 거래에서의 지침에 중점을 두는 반면 우리나라의 전자거래기본법과 전자서명법은 정부의 역할을 중시한 것이 특징이다. 통일전자거래법에는 우리나라와 동일하게 전자기록, 전자서명등에 대한 정의가 있고 자동화된 거래, 컴퓨터 프로그램, 전자대리인, 정보, 정보처리시스템, 보안절차등 우리나라에 없는 정의가 들어 있다. 제16조에서 규정하는 양도가능한 기록(Transferable Record)도 우리나라에 없는 특이한 개념이다. 여기서 명시하고 있는 전자어음은 법제정이 있기도 진에 우리나라에 도입하려고 하는 개념으로 금융산업의 지각변동을 예고하는 조항이기도 하다. 우리나라는 사이버몰, 인증기관을 강조하고 소비자의 보호지침을 도입한 것이 미국과 다른 특징이다. 미국의 경우는 각종 분쟁의 경우 법원의 해석을 강조하여 전통적인 사법부우위의 모습을 보여주고 있다. 향후 예상되는 각종 분쟁에 대한 미국의 판례연구는 한국에서도 신속하게 연구되어 적용되어야 할 분야라 하겠다.

• 무역상무연구
• /
• 제73권
• /
• pp.21-42
• /
• 2017

A BPO is an irrevocable undertaking given by an Obligor Bank acting on behalf of the buyer to a Recipient Bank acting on behalf of the seller to pay on satisfaction of certain specified conditions. There have been a Trade Service Utility(TSU), a Transaction Matching Application(TMA) and ISO 20022 TSMT messages that core electronic technical systems linked to the BPO. As a key usefulness of the BPO, it brings to market an alternative means of satisfying the risk mitigation, financing and information management needs of banks and businesses engaged in trade. That is, corporates can spread the risk among multiple Obligor Banks by requesting multiple BPOs for the same trade transaction not to exceed the value of the initial transaction. Since the BPO replaces the manual document checking process with the electronic matching of data, buyers and sellers will benefit from significantly increased accuracy and objectivity. By using BPO, buyers and sellers can become trusted counterparties by demonstrating reliability and giving sellers the assurance of being paid on time as per the payment terms and conditions agreed. When compared to L/C, the BPO requires submission of data only. This requires a change of practice of those accustomed to dealing with physical documentation. The beneficiary of a BPO is the Recipient Bank which is always the Seller's Bank. In the meantime, banks will need to agree between themselves that the URBPO 750 will be applied, modified or excluded. However, customers of banks will face low transactional costs due to a fully automated process and standardized data formats used in international trade documentation. There is no amendment and cancellation rules in the URBPO 750s, but only have Assignment of Proceeds rules. As a result, the BPO is likely to enable new business opportunities, lower costs, fast transaction process and strengthen key customer relationships.

• 한국경영과학회:학술대회논문집
• /
• 한국경영과학회/대한산업공학회 2003년도 춘계공동학술대회
• /
• pp.772-779
• /
• 2003

In the real world, business processes are very complex and composed or heterogeneous business activities. As the advent or the Web services enables business processes to be Integrated and to be automated. It makes enterprises integrate heterogeneous business processes from different business partners as well as their internal business processes. To support recent trends in integration of business processes, BPEL, WS-E and WS-T specipications have been established since 2002 WS-C and WS-T describe the reliable business environment including compensation (undo) or completed business processes. The compensation of business processes is a basic requirement for automation or business processes among business partners. Systems implementing these specifications, however, very rarely exist. It's not only because those specifications are developed recently. but also because they are not perfect yet. In this paper, a new business process transaction management, which complements the deficiency of WS-E and WS-T, is suggested. Furthermore, the new approach proposes the business logics for supervisory coordinators which manage serial and parallel business gates The modification or traditional WS-T specification and the simplification or WS-E specification make business processes managed effectively.