• 대한전기학회논문지:시스템및제어부문D
• /
• 제51권4호
• /
• pp.152-159
• /
• 2002

Fitness sharing that shares fitness if calculated distance between individuals is smaller than sharing radius is one of the representative speciation methods and can complement evolutionary algorithm which converges one solution. Recently, there are many researches on designing neural network architecture using evolutionary algorithm but most of them use only the fittest solution in the last generation. In this paper, we elaborate generating diverse neural networks using fitness sharing and combing them to compute outputs then, propose calculating distance between individuals using modified Kullback-Leibler entropy for improvement of fitness sharing performance. In the experiment of Australian credit card assessment, breast cancer, and diabetes in UCI database, proposed method performs better than not only simple average output or Pearson Correlation but also previous published methods.

• 한국GIS학회:학술대회논문집
• /
• 한국GIS학회 2003년도 추계학술대회논문집
• /
• pp.9-14
• /
• 2003

The major reason that spatial data warehousing has attracted a great deal of attention in business GIS in recent years is due to the wide availability of huge amounts of spatial data and the imminent need for turning such data into useful geographic information. Therefore, this research has been focused on designing and implementing the pilot tested system for spatial decision making. The purpose of the system is to predict targeted marketing area by discriminating the customers by using both transaction quantity and the number of customer using credit card in department store. Focused on the analysis methodology, the case study is aiming to use GIS and clustering for knowledge discovery. The system is a key section of the research of multi-dimensional and spatio-temporal analysis in the internet environment.

• 통상정보연구
• /
• 제7권1호
• /
• pp.75-94
• /
• 2005

With most major full service banks having launched transactional Internet banking, attention is shifting to the realities of managing the Internet channel as a profitable component of an overall delivery strategy. In addition to examining Case Study and Internet Bank Model. Services of Internet Banking available through the Internet are as follows. 1. credit card loans, personal loans. 2. high-yield financial products. 3. insurance products. 4. securities products. 5. Case study of Foreign Internet Banking(ING, BNP, HSNC, City Bank). The study reviewed fields, including financial services, customer service, Website formation and design, convenience of use and system safety, Internet Banking Model, and many related areas. Internet Banking earned high marks in most fields. This study review focuses on the following: Understanding and meeting consumer expectations for us ability, site performance and functionality. Integrating the Internet channel into overall marketing, product delivery and customer service strategies. Strategies to increase customer satisfaction with Internet Banking and to attract new Internet bankers. therefore this study review activity model concretion of Internet Banking Model and Case Study.

• 한국생활과학회지
• /
• 제14권6호
• /
• pp.973-983
• /
• 2005

This study examined the practice of the impulsive purchase and the factors influencing it among college consumers. The data were collected from 481 college students in Busan and Gyeongnam areas through a self-administered questionnaire. Frequencies, Cronbach's alpha, factor analysis, Pearson's correlation analysis, and regression analysis were conducted by SPSS Windows. The results from this study were as follows: First, the level of the impulsive purchase of college consumers was not that high. Second, as a result of regression analysis for the group of the high level of impulsive purchase, the allowance amount, price-discount factor, credit card use, psychological factor, and attitudes towards money (ostentatious way, and symbol of success), affected the impulsive purchase. These results suggest that consumer education regarding values about money and strategies to resist impulsive purchase should be conducted at home and in school.

• International Journal of Computer Science & Network Security
• /
• 제21권6호
• /
• pp.200-206
• /
• 2021

Fraud in e-commerce transaction increased in the last decade especially with the increasing number of online stores and the lockdown that forced more people to pay for services and groceries online using their credit card. Several machine learning methods were proposed to detect fraudulent transaction. Neural networks showed promising results, but it has some few drawbacks that can be overcome using optimization methods. There are two categories of learning optimization methods, first-order methods which utilizes gradient information to construct the next training iteration whereas, and second-order methods which derivatives use Hessian to calculate the iteration based on the optimization trajectory. There also some training refinements procedures that aims to potentially enhance the original accuracy while possibly reduce the model size. This paper investigate the performance of several NN models in detecting fraud in e-commerce transaction. The backpropagation model which is classified as first learning algorithm achieved the best accuracy 96% among all the models.

• International Journal of Internet, Broadcasting and Communication
• /
• 제8권1호
• /
• pp.7-18
• /
• 2016

Several authentication methods have been developed to make use of tokens in the mobile networks and smart payment systems. Token used in smart payment system is genearated in place of Primary Account Number. The use of token in each payment transaction is advantageous because the token authentication prevents enemy from intercepting credit card number over the network. Existing token authentication methods work together with the cryptogram, which is computed using the shared key that is provisioned by the token service provider. Long lifetime and repeated use of shared key cause potential brawback related to its vulnerability against the brute-force attack. This paper proposes a per-transaction shared key mechanism, where the per-transaction key is agreed between the mobile device and token service provider for each smart payment transaction. From server viewpoint, per-transaction key list is easy to handle because the per-transaction key has short lifetime below a couple of seconds and the server does not need to maintain the state for the mobile device. We analyze the optimum size of the per-transaction shared key which satisfy the requirements for transaction latency and security strength for secure payment transactions.

• 지능정보연구
• /
• 제20권4호
• /
• pp.59-87
• /
• 2014

결제서비스에 대한 기존의 연구는 결제서비스의 채택요인 또는 지속적인 사용에 영향을 미치는 요인 등 행동이론을 중심으로 진행되어 왔다. 이러한 요인들이 미치는 영향에 대한 결과는 결제서비스의 종류에 따라 또는 연구 지역에 따라 상이하게 나타나고 있다. 본 연구는 결제 서비스의 종류나 문화등의 변수에 관계없이 새로운 결제 서비스가 성공할 수 있는 일반적인 요인이 무엇인지에 대한 의문에서 시작하게 되었다. 기존 연구에서 중요한 영향을 미친다고 제시한 채택요인들은 실제 결제사례의 결과에 비추어 보면 기존 연구에서 주장한 바와 일치하지 않는 경우를 볼 수 있다. 이러한 이론과 현실사이의 괴리를 발견하고 새로운 결제서비스가 성공하기 위한 근본적이고 결정적인 요인이 무엇인지에 대해 제시하고 사례연구를 통해 가설을 입증하고자 하는 것이 본 연구의 목적이다. 따라서 본 연구는 새로운 결제서비스가 성공하기 위해서는 기존 결제서비스의 비고객에게 이들이 결제할 수 있는 수단을 제공함으로써 새로운 결제 시장을 창출해야 함을 주장한다. 이를 위해 성공한 결제사례인 신용카드, 휴대폰 소액결제, PayPal, Square을 채택하였으며, 기존 결제서비스의 비고객을 3개의 계층으로 분류하여 분석하였다. 그리고 새로운 결제서비스가 어떠한 계층을 타겟으로 하였으며 이들에게 어떠한 결제수단을 제공하여 새로운 시장을 창출하였는지 제시한다. 사례 분석 결과, 성공 사례 모두 본 연구의 가설을 지지하는 것으로 나타났다. 따라서 새로운 결제서비스는 결국 기존의 결제수단으로 거래를 할 수 없었던 이들이 결제를 할 수 있도록 함으로써 성공할 수 있다는 가설을 입증하였다. 모바일 결제서비스가 아직 대중화되지 못한 원인을 본 가설에 비추어 분석해 보면 보면, 기존의 결제 인프라를 이용할 수 있는 바코드, QR코드 기반의 모바일 결제 서비스뿐만 아니라 NFC, BLE, 음파 등의 새로운 기술이 적용된 모바일 결제 서비스가 출시되는 등 새로운 시도가 계속되고 있다. 또한 모바일 월렛은 사용자들이 소지하고 있는 카드정보를 스마트폰에 저장하여 지갑 없이도 결제가 가능하며, 쿠폰 제공, 적립카드 관리, 신분증을 저장하는 등의 다양한 부가적인 기능을 제공하고 있어 성공할 것이라는 전망이 대두되고 있다. 하지만 이러한 서비스들은 본 연구 관점에서 보자면 기존 결제서비스의 비고객이(기존 결제수단을 이용할 수 없었던 사용자) 거래할 수 있는 새로운 결제 수단을 제공해 주지 못하고 있기 때문에 결국 초기사용자에게만 채택될 뿐 대중화되는데 한계가 있을 것으로 예상된다. 반면, 새로운 모바일 결제서비스의 성공사례 중 하나인 PaybyPhone은 기존 코인주차 결제서비스의 비고객인 현금 미소지 고객에게 스마트폰을 이용한 새로운 결제수단을 제공함으로써 새로운 주차 결제 시장을 창출하였으며 현재 미국뿐만 아니라 유럽시장까지 진출하는 등 급성장하고 있다. 결론적으로, 많은 이해관계자들이 모바일 결제시장을 선점하기 위해 다양한 형태의 모바일 결제 서비스를 출시하고 있지만 캐즘을 뛰어넘어 주류 시장에 성공적으로 정착할 수 있느냐는 결국 기존 결제서비스의 비고객군에게 그들이 필요로 하는 새로운 결제수단을 제공하는지의 여부에 달려있다고 볼 수 있다. 따라서 모바일 결제 서비스의 기획자나 매니저들은 서비스 기획 시 기존 결제서비스의 비고객군은 누구인가? 그들은 어떠한 결제수단을 원하는가?를 먼저 고려해야 한다. 본 연구는 새로운 결제서비스가 성공하는데 미치는 요인에 대한 가설을 검증하기 위해 4개의 성공사례를 선택하였으며 각 사례에 동일한 가설을 검증하는 '반복연구논리'를 적용하였다. 본 가설을 더욱 공고히 하기 위해 사례연구방법론에서 제시하고 있는 경쟁가설을 포함한 후속 사례연구가 진행되어야 할 것이다.

• 가족자원경영과 정책
• /
• 제10권3호
• /
• pp.23-43
• /
• 2006

The purpose of this study is to identify the types of consumer problems involving the use of online game services in Korea. This study analyzed the consumer damage cases by age and type that were reported to the Korea Consumer Protection Board (KCPB). The results of this study are as follows: The online game complaints varied by age. In the case of children, most complaints involved phone charging online game fees without parents' consent. There were similarities in the victimization of teens and children according to reason for complaint, handling result and type of damage. However, based on method of charging, mobile phone and credit card payments were causes in the case of teen victimization. Meanwhile, complaints among adults showed more complexity than the previous two groups such as cases relating to online game companies illegally seizing game items and game accounts, and charging fees for unused services caused by identity theft. By type of charging method, most complaints involved phone and mobile phone payments. According to the analysis result of online game victims, the monetary damage of consumers did not differ among socio-demographic variables such as age, residence, gender and occupation but there was difference in the method of charging. In the handling period, differences could be seen by age, year, gender and type of damage.

• International journal of advanced smart convergence
• /
• 제8권2호
• /
• pp.102-108
• /
• 2019

Big data was associated with three key concepts, volume, variety, and velocity. Securities and investment services produce and store a large data of text/numbers. They have also the most data per company on the average in the US. Gartner found that the demand for big data in finance was 25%, which was the highest. Therefore securities and investment companies produce the largest data such as text/numbers, and have the highest demand. And insurance companies and credit card companies are using big data more actively than banking companies in Korea. Researches on the use of big data in securities and investment companies have been found to be insignificant. We surveyed 22 major securities and investment companies in Korea for activating big data. We can see they actively use AI for investment recommend. As for big data of securities and investment companies, we studied open API. Of the major 22 securities and investment companies, only six securities and investment companies are offering open APIs. The user OS is 100% Windows, and the language used is mainly VB, C#, MFC, and Excel provided by Windows. There is a difficulty in real-time analysis and decision making since developers cannot receive data directly using Hadoop, the big data platform. Development manuals are mainly provided on the Web, and only three companies provide as files. The development documentation for the file format is more convenient than web type. In order to activate big data in the securities and investment fields, we found that they should support Linux, and Java, Python, easy-to-view development manuals, videos such as YouTube.

• 한국컴퓨터정보학회논문지
• /
• 제24권8호
• /
• pp.77-85
• /
• 2019

In this paper, we propose a "Safe Web Using Scrapable Headless Browse" Because in a network separation environment for security, It does not allow the Internet. The reason is to physically block malicious code. Many accidents occurred, including the 3.20 hacking incident, personal information leakage at credit card companies, and the leakage of personal information at "Interpark"(Internet shopping mall). As a result, the separation of the network separate the Internet network from the internal network, that was made mandatory for public institutions, and the policy-introduction institution for network separation was expanded to the government, local governments and the financial sector. In terms of information security, network separation is an effective defense system. Because building a network that is not attacked from the outside, internal information can be kept safe. therefore, "the separation of the network" is inefficient. because it is important to use the Internet's information to search for it and to use it as data directly inside. Using a capture method using a Headless Web browser can solve these conflicting problems. We would like to suggest a way to protect both safety and efficiency.