• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2021.05a
• /
• pp.100-103
• /
• 2021

Most companies study why employees resign their jobs to prevent the outflow of excellent human resources. To obtain the data needed for the study, employees are interviewed or surveyed before resignation. However, it is difficult to get accurate results because employees do not want to express their opinions that may be disadvantageous to working in a survey. Meanwhile, according to the data released by the Korea Labor Institute, the greater the difference between the minimum level of education required by companies and the level of employees' academic background, the greater the tendency to resign jobs. Therefore, based on these data, in this study, we would like to predict whether employees will leave the company based on data such as major, education level and company type. We generate four kinds of resignation prediction models using Decision Tree, XGBoost, kNN and SVM, and compared their respective performance. As a result, we could identify various factors that were not covered in previous study. It is expected that the resignation prediction model help companies recognize employees who intend to leave the company in advance.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.27 no.4
• /
• pp.763-773
• /
• 2017

In recent years, personal information leakage and technology leakage accidents are frequently occurring. According to the survey, the most important part of this spill is the 'insider' within the organization, and the leakage of technology by insiders is considered to be an increasingly important issue because it causes huge damage to the organization. In this paper, we try to learn the normal behavior of employees using machine learning to prevent insider threats, and to investigate how to detect abnormal behavior. Experiments on the detection of abnormal behavior by implementing an Autoencoder composed of Recurrent Neural Network suitable for learning time series data among the neural network models were conducted and the validity of this method was verified.

• Journal of Korea Multimedia Society
• /
• v.14 no.2
• /
• pp.228-237
• /
• 2011

The risk of leakage or theft of critical data which is stored in database is increasing in accordance with evolution of information security paradigm. At the same time, needs for database security have been on the rapid increase due to endless leakage of massive personal information. The existing technology for prevention of internal information leakage possesses the technical limitation to achieve security goal completely, because the passive control method including a certain security policy, which allows the only authorized person to access to DBMS, may have a limitation. Hence in this study, we propose Query Signature System which signatures the queries accessing to the critical information by interrupting and passing them. Furthermore this system can apply a constant security policy to organization and protect database system aggressively by restricting critical query of database.

• KIPS Transactions on Computer and Communication Systems
• /
• v.5 no.2
• /
• pp.33-40
• /
• 2016

Due to the growth of social network systems (SNS), big data are realized and Hadoop was developed as a distributed platform for analyzing big data. Enterprises analyze data containing users' sensitive information by using Hadoop and utilize them for marketing. Therefore, researches on data encryption have been done to protect the leakage of sensitive data stored in Hadoop. However, the existing researches support only the AES encryption algorithm, the international standard of data encryption. Meanwhile, Korean government choose ARIA algorithm as a standard data encryption one. In this paper, we propose a HDFS data encryption scheme using ARIA algorithms on Hadoop. First, the proposed scheme provide a HDFS block splitting component which performs ARIA encryption and decryption under the distributed computing environment of Hadoop. Second, the proposed scheme also provide a variable-length data processing component which performs encryption and decryption by adding dummy data, in case when the last block of data does not contains 128 bit data. Finally, we show from performance analysis that our proposed scheme can be effectively used for both text string processing applications and science data analysis applications.

• Journal of Wetlands Research
• /
• v.23 no.1
• /
• pp.67-73
• /
• 2021

As the impermeable area increases due to urbanization and industrialization, the influence of non-point pollutants caused by rainfall runoff on the water system is increasing. In the past, the best management practices(BMP) were used a lot to manage non-point pollutants, but recently, technology that naturally treats them through LID (Low Impact Development) technology is widely used. In this study, various rainfall events were simulated through the SWMM model based on the data of rainfall monitoring in bioretention among natural facilities. The characteristic of LID modeling research is that it is difficult to build accurate modeling data with short-term data because real data is the result obtained through natural facilities, and it is difficult to implement an accurate model. In this study, the data monitored for 3 years It is significant in that it has built a precise model. The actual data monitored a total of 18 times was simulated, and the inflow and outflow and the removal efficiency of five pollutants were simulated. As a result of performing the performance evaluation, most of the 7 items showed excellent indicators, and the TN and TP showed relatively low simulation performance. In the future, it is expected that Korea will introduce an integrated water management system in which the water supply system and the sewage system are substantially integrated and operated. Therefore, the results of this study are considered to play an important role in the initial stage of rainfall management in the future integrated water management system, and the extent of rainfall runoff reduction and pollutant reduction in the expected installation area can be predicted in advance. This is expected to prevent overdesign of bioretention.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.39C no.1
• /
• pp.1-8
• /
• 2014

This paper proposed a method that prevents data leakage and modulation and detects an unauthorized device by using AP and mobile device. Most of existing method for detecting and preventing an unauthorized device are similar to type of Evil Twin. However, in previous studies can not resolve many security accident and have the problem to cope with effectively security accident on various network. In order to solve these problem, companies prepare security accident through the varies policy and guideline. but It is hard to prevent all security accident because it is consistently increasing everyday. This paper suggests technique of detecting and preventing an unauthorized device using Context Awareness and Mobile Device Management. Firstly, when mobile device go into internal network, server distinguish access permission and authorization of mobile device using acquiring the conetxt information of mobile device. By using this result, server applies the appropriate management technique to the mobile device for leakage and accident of internal network.

• Convergence Security Journal
• /
• v.18 no.5_2
• /
• pp.21-27
• /
• 2018

Most of the Internet users in Korea are issued certificates and use them for various tasks. For this reason, it is recommended that accredited certification authorities and security related companies and use public certificates on USB memory and portable storage devices rather than on the user's desktop. Despite these efforts, the hacking of the certificate has been continuously occurring and the financial damage has been continuing. Also, for security reasons, our military has disabled USB to general military users. Therefore, this study proposes a two-factor method using the unique information of the USB memory and the PC which is owned by the user, and suggests a method of managing the private key file secure to the general user. Furthermore, it will be applied to national defense to contribute to the prevention of important data and prevention of access by unauthorized persons.

• Proceedings of the Korea Society for Industrial Systems Conference
• /
• 2009.05a
• /
• pp.20-26
• /
• 2009

본 연구는 건물 내부에서 외부로 외부에서 내부로 출입하는 출입자를 정확히 파악하기 위하여 보안이 강화된 개별인증 출입보안시스템을 구축하는데 있다. 시스템 구현을 위해서 RFID 기술, 지문인식 기술을 적용한 통합인증단말기와 무게를 측정하는 전자저울을 활용 하였다. 개별 인증이 가능한 출입 인증 보안시스템을 구현한 결과를 요약하면, 첫째, 내부에서 외부로 나오는 경우와 외부에서 내부로 들어가는 경우를 분리한 이중 출입방식으로 외부인이 내부에서 나오는 사람의 인증만으로 쉽게 건물 안으로 진입하는 보안상의 문제점을 해결 하였다. 둘째, 건물 내부 물품을 외부로 유출할 시 무게 차이 때문에 출입문이 차단되어 관리자의 허락 없이는 물품을 밖으로 유출할 수 없게 되어 도난을 방지문제를 해결 하였다. 셋째, 몸무게 정보를 이용하여 출입자들이 동시에 출입문을 통과하지 못하도록 제한하여 모든 출입자의 로그정보를 정확하게 저장할 수 있도록 하였다. 넷째, 데이터베이스에 저장된 출입로그 데이터를 출퇴근 관리프로그램에 지원하여 급여 계산 및 인사 업무에 활용할 수 있도록 하였다. 다섯째, 건물에 화재가 나거나 폭탄 테러와 같은 위험 발생 시 출입자를 정확하게 파악할 수 있게 되어 용의자 파악이 용이하도록 하였다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2008.05a
• /
• pp.1112-1115
• /
• 2008

인터넷으로 대표되는 정보통신망 및 컴퓨터를 이용한 개인 정보 수집과 활용이 일반화됨에 따라 수집된 개인정보의 불법적인 접근 유출 사례가 증가하고 있다. 현재의 개인정보 이용 환경은 데이터 접근 시 사용자의 질의 내용과 그에 대한 결과가 그대로 노출되어 사용자의 프라이버시를 침해하는 문제를 안고 있다. 본 논문에서는 데이터베이스에서 개인정보보호를 위해 접근제어 정책 기반 쿼리 처리시스템인 보안 게이트웨이를 설계하고 구현한다. 이 시스템은 클라이언트가 TDS 프로토콜을 이용하여 DBMS에 접근해 정보를 요청할 때 보안 정책을 반영함으로써 단순한 차단은 물론 변환된 쿼리 응답을 한다. 본 시스템은 불법적인 접근에 대한 제어는 물론이고, 정당한 인증자의 실수나 고의적인 개인정보 유출로 인한 경제적, 사회적 손실을 방지할 수 있다. 또한 주민등록번호 등 보안 대상 정보를 제외한 기타 정보에 대한 접근을 허용함으로써 데이터베이스 가용성을 보장한다.

• Journal of Convergence Society for SMB
• /
• v.3 no.1
• /
• pp.43-48
• /
• 2013

With the development of computer technology, and have diversified technical system attacks evolve more. Computer hardware and software has evolved more and more, performance is improved, but the basic principle of operation does not change much, it is a problem. In general, the application is placed in a running state, the program data is placed in memory. Remains in memory for efficiency of operation of the operating system, we analyze memory and memory hacking, these data will have access to data. Since a large damage occurs key, such as certificates personal information, encrypted flows out, measures should be provided by it. In this content, I want to discuss the issues and work around memory hacking.