http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Hadoop Parameter 변경을 통한 MapReduce 성능 최적화
지경엽(Keungyeup Ji),권영미(Youngmi Kwon) 한국정보기술학회 2020 Proceedings of KIIT Conference Vol.2020 No.10
본 논문에서는 디스크 처리 기반으로 대용량 데이터 처리를 하는 Hadoop 성능 향상을 위해서 Hadoop 환경의 MapReduce 관련 복제개수와 bock size parameter 조정을 통해서 최적화 환경을 파악하고자 하는 것이 목적이다. 그러기 위해서는 실험대상의 parameter 들을 최적화하여 Hadoop jar 실행문을 이용하여 Map, Reduce 방식으로 세 개의 directory에 분산 저장함으로써, job을 수행하는데 가장 성능이 좋은 최적화된 환경을 찾고자 하는 것이다. In this paper, in order to improve the performance of Hadoop, which processes large amounts of data based on disk processing, the purpose is to grasp the optimization environment by adjusting the number of replication and bock size parameters related to MapReduce in the Hadoop environment. To do this, we want to find the optimized environment with the best performance to perform the job by optimizing the parameters of the experiment and storing them in three directories using Map and Reduce method by Hadoop jar execution statement. Using this kind of method, it is to find the best performance and an optimized environment to perform the main job.
Hadoop Parameter 조정을 통한 Hadoop MapReduce 성능 최적화 분석 연구
지경엽(Keungyeup Ji),권영미(Youngmi Kwon) 한국정보기술학회 2021 한국정보기술학회논문지 Vol.19 No.6
The purpose of this paper is to understand the performance optimization environment through parameter adjustment of the Hadoop environment as the most efficient way to solve the inefficiency of speed, which is the disadvantage of the Hadoop framework that processes large amounts of data based on distributed processing and virtualization. To this end, we derived optimization of the HDFS and MapReduce performance improvement that support Hadoop. In this paper, through this test, in order to utilize the Hadoop framework in an optimal state under a given resource condition, it is considered efficient to perform data processing after deriving an appropriate configuration environment through Hadoop parameter adjustment.
멀티노미얼나이브베이즈 기법의 정교화를 통한 악성 메일 필터링 시스템 구현
지경엽(Keungyeup Ji),권영미(Youngmi Kwon) 한국정보기술학회 2023 한국정보기술학회논문지 Vol.21 No.7
In a situation that malicious e-mails are seriously increased, in order to increase accuracy of malicious mail machine learning methods of three kinds instead of message rule-based methods were applied to this paper. The main goal is to prove that the multinomial Naive Bayes technique of a Naive Bayes algorithm has better prediction accuracy for malicious emails and takes less processing time than other two kinds of machine learning techniques. To prove this, according to the experimental result consisting of 1,454,489 the spam prediction error rate was 8% for the MultinomialNB algorithm 42% for the SVM algorithm and 20% for the LR algorithm. In conclusion, I suggest that implementing a malicious mail filtering system by applying the Laplace smoothing technology and good alpha parameter value based on the MultinomualNB algorithm is an effective method.