• 한국정보교육학회:학술대회논문집
• /
• 2021.08a
• /
• pp.295-301
• /
• 2021

This study tried to examine the necessity of AI literacy education to increase the effectiveness of artificial intelligence education ahead of the revision of the next revised curriculum. To this end, AI modeling classes were conducted for high school students and the necessity, content, and training period of AI literacy perceived by students in AI education were investigated through a questionnaire. The results showed that they generally agreed on the need for data utilization and data preprocessing in the AI class, and in the course of the AI class, there were many cases of difficulties due to lack of basic competencies for database use. In particular, it was observed that the understanding of the file structure for data analysis was insufficient and the understanding of the data storage format for data analysis was low. In order to overcome this part, the necessity of prior education for data processing was recognized, and there were many opinions that it is generally appropriate to go to high school at that time. As for the content elements of AI literacy, it was found that there were high demands on the content of data visualization along with data transformation, including data creation and deletion.

• Journal of Information Technology Applications and Management
• /
• v.11 no.2
• /
• pp.205-214
• /
• 2004

Internet evolves into the huge network with new nodes inserted or deleted depending on specific situations. A new model of network topology is needed in order to analyze time-varying Internet more realistically and effectively. In this study the non-monotony models are proposed which can describe topological changes of Internet such as node insertion and deletion, and can be used for predicting its future topology. Simulation is performed to analyze the topology generated by our model. Simulation results show that our proposed model conform the power law of realistic Internet better than conventional ones. The non-monotony model can be utilized for designing Internet protocols and networks with better security.

• Journal of Korean Institute of Industrial Engineers
• /
• v.20 no.4
• /
• pp.73-84
• /
• 1994

Output analysis is one of the most important fields of simulation to achieve the accurate simulation results. This study shows how to analyze simulation output data in the steady state using Procrustes analysis technique which has not been used in the field of simulation yet. In this paper Procrustes analysis method is used to perform the analysis of simulation output efficiently and effectively by applying the improved version of the method. The experiments are conducted using M/M/1 queueing simulation model. The results obtained by Procrustes analysis method show better estimates for average waiting times and average queue lengths which are closer to true values and narrower confidence intervals than when replication-deletion method is used. Also it requires the smaller number of simulation runs.

• Journal of the Korean Statistical Society
• /
• v.29 no.3
• /
• pp.285-294
• /
• 2000

In analyzing bivariate date the sample correlation coefficient is often used, and it is quite sensitive to one or few isolated cases. In this article we derive a formula for the effect of $textsc{k}$ observations on the samples correlation coefficient by the deletion method. To give a reference value for the isolated cases the asymptotic distribution fo the formula is derived. Also, we give some interpretations on several types of isolated cases and an example based on a real data set.

• The Korean Journal of Physiology and Pharmacology
• /
• v.2 no.5
• /
• pp.645-650
• /
• 1998

To assess the relationship between angiotensin-converting enzyme (ACE) gene polymorphism and myocardial infarction in Koreans, we recruited 112 healthy, unrelated subjects (mean age 53.4 years) and 104 myocardial infarction survivors (mean age 54.2 years) of both sexes. An insertion/deletion (I/D) polymorphism of the ACE gene was typed by polymerase chain reaction. The I allelic frequency of ACE gene in Korean subjects was irrelavant to myocardial infarction (patients, 65 control subjects 66%), as was true with the D allele. When compared with other populations, the frequency of D allele in Koreans (0.34) was lower than that in Caucasians, and was close to that of other Oriental populations. The data suggest that the ACE gene polymorphism is not an independent genetic risk factor for myocardial infarction in Koreans.

• The Journal of Information Systems
• /
• v.23 no.3
• /
• pp.127-142
• /
• 2014

Plagiarism occurs when the content is copied without permission or citation, and the problem of plagiarism has rapidly increased because of the digital era of resources available on the World Wide Web. An important task in plagiarism detection is measuring and determining similar text portions between a given pair of documents. One of the main difficulties of this task is that not all similar text fragments are examples of plagiarism, since thematic coincidences also tend to produce portions of similar text. In order to handle this problem, this paper proposed association analysis in data mining to detect plagiarism. This method is able to detect common actions performed by plagiarists such as word deletion, insertion and transposition, allowing to obtain plausible portions of plagiarized text. Experimental results employing an unsupervised document classification strategy showed that the proposed method outperformed traditionally used approaches.

• Animal cells and systems
• /
• v.7 no.3
• /
• pp.249-253
• /
• 2003

Abnormalities in fibrinolysis system is associated with risk of hypertension. In this report, the Alu repeat insertion/deletion (I/D) polymorphism of tissue plasminogen activator (t-PA) and the Hind III RFLP of plasminogen activator inhibitor-1 (PAI-1) genes were investigated in 115 normotensives and 83 patients with hypertension, and their association with anthropometrical data and plasma biochemical parameters were analyzed. There were no significant differences in the gene frequencies of the two candidate genes between normotensives and hypertensives, respectively. Our results indicate lack of associations between the two polymorph isms in t-PA and PAI-1 genes and risk of hypertension in the population under study. However, the Hind III RFLP of PAI-1 gene was significantly associated with plasma glucose level, suggesting its role in glucose metabolism. It needs to be tested whether this RFLP of PAI-1 gene is associated with insulin resistance syndrome or non-insulin dependent diabetes mellitus (NIDDM) in the Korean population.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2018.10a
• /
• pp.429-430
• /
• 2018

The introduction of mobile devices increased the need to apply limitations such as mobile devices' memory, speed, energy, and bandwidth on the designs of searching images. There is a demand to reduce such limitations on searching images on the mobile phone. Hence, this paper proposes a design that adds tags on pictures to manage the images in mobile environment, allowing efficient searches and deletion of duplicate files based on the similarities of the images. The proposed method does not compromise its efficiency by increasing costs; it also reduces the volume of data needed for mobile devices.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2014.07a
• /
• pp.241-244
• /
• 2014

본 논문에서는 HDFS(Hadoop Distributed File System)에서 문제되고 있는 복제정책의 개선을 위해 HMM(Hidden Markov Model)을 이용한 동적 데이터 복제본 삭제 전략을 제안한다. HDFS는 대용량 데이터를 효과적으로 처리할 수 있는 분산 파일 시스템으로 높은 Fault-Tolerance를 제공하며, 데이터의 접근에 높은 처리량을 제공하여 대용량 데이터 집합을 갖는 응용 프로그램에 최적화 되어있는 장점을 가지고 있다. 하지만 HDFS 에서의 복제 메커니즘은 시스템의 안정성과 성능을 향상시키지만, 추가 블록 복제본이 많은 디스크 공간을 차지하여 유지보수 비용 또한 증가하게 된다. 본 논문에서는 HMM과 최상의 상태 순서를 찾는 알고리즘인 Viterbi Algorithm을 이용하여 불필요한 데이터 복제본을 탐색하고, 탐색된 복제본의 삭제를 통하여 HDFS의 디스크 공간과 유지보수 비용을 절약 할 수 있는 전략을 제안한다.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• v.8 no.2
• /
• pp.121-125
• /
• 2008

It is essential to maintain hierarchical information properly for efficient XML query processing. Well known approach to represent hierarchical information of XML tree is assigning a specific node number to each node of XML tree. Insertion and deletion of XML node can occur at any position in a dynamic XML tree. A dynamic numbering scheme allows us to add nodes to or delete nodes from an XML tree without relabeling or with relabeling only a few existing nodes of XML tree while executing XML query efficiently. According to W3C XQuery update facility specifications a node can be added as first or last child of the existing node in XML tree. Generating new number for last child requires referencing the number of previous last child. Getting the number of last child is very costly with previous approaches. We have developed a new dynamic numbering scheme PSN which is very effective for insertion of a node as last child. Our approach reduces the time to find last child dramatically by removing sorting of children.