• Journal of Information Processing Systems
• /
• v.18 no.3
• /
• pp.302-310
• /
• 2022

Efficient use of limited blood products is becoming very important in terms of socioeconomic status and patient recovery. To predict the appropriateness of patient-specific transfusions for the intensive care unit (ICU) patients who require real-time monitoring, we evaluated a model to predict the possibility of transfusion dynamically by using the Medical Information Mart for Intensive Care III (MIMIC-III), an ICU admission record at Harvard Medical School. In this study, we developed an explainable machine learning to predict the possibility of red blood cell transfusion for major medical diseases in the ICU. Target disease groups that received packed red blood cell transfusions at high frequency were selected and 16,222 patients were finally extracted. The prediction model achieved an area under the ROC curve of 0.9070 and an F1-score of 0.8166 (LightGBM). To explain the performance of the machine learning model, feature importance analysis and a partial dependence plot were used. The results of our study can be used as basic data for recommendations related to the adequacy of blood transfusions and are expected to ultimately contribute to the recovery of patients and prevention of excessive consumption of blood products.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2021.05a
• /
• pp.175-177
• /
• 2021

The parking area for the disabled is a parking facility for the pedestrian disabled and is a parking space for securing pedestrian safety passage for the disabled. However, due to the lack of social awareness of areas for the disabled, the use of parking areas is restricted, and violations such as illegal parking and obstruction of parking are increasing every year. Therefore, in this study, we propose a system to crack down on illegal parking in handicapped parking areas using the YOLOv5 model, a deep learning object recognition model to improve parking interference within parking spaces.

• Educational Technology International
• /
• v.14 no.1
• /
• pp.75-108
• /
• 2013

The purpose of this study is to find a way to improve digital textbooks for self-regulated learning by applying cognitive scaffolding designs to an elementary math digital textbook and examining the effectiveness of the system. Hence this study was conducted in two steps. First, a framework for scaffolding design was devised by examining the problems and difficulties students encounter when using a mathematics digital textbook. Second, after the digital textbook was revised by applying the scaffolding design frameworks, the effectiveness of the scaffolding framework was examined by comparing students' achievement levels in an experimental group and that of students in a control group. Seventy fifth-graders participated in this study. Students were divided into two groups: an experimental group and a control group. The students in the experimental group studied with the revised version of the digital textbook and the students in the control group studied with the original version of the digital textbook. The students received a pretest before the experiment. After the experiment, they took an achievement test and completed a usability questionnaire. The data were analyzed by ANCOVA with the SPSS Windows version. The results revealed that the students who used the revised program (to which design strategies for scaffolding were applied) showed higher levels of achievement than those who used the original version. In addition, students in the experimental group generally showed higher scores on the usability survey, which consisted of four sub-categories such as 'effectiveness', 'efficiency', 'satisfaction', and 'learnability'. There was a statistically significant effect on 'efficiency'. These results implied that scaffolding strategies were effective for mathematics learning through the use of an elementary digital textbook.

• Educational Technology International
• /
• v.7 no.1
• /
• pp.1-20
• /
• 2006

This paper describes practical research on the improvement of teaching and learning process by integrating Web Bulletin Board (WBB) and mobile phone. This paper addresses three topics; A) the interactive lecture with topics-based discussions using the Web Bulletin Board (WBB) as a tool for assisting discussion, B) the introduction of peer evaluation among students to develop their problem-solving and cognitive skills, C) the use of mobile phones for promoting interactive lectures, keeping class attendance, conducting assignments, and providing notices for the next class. Results indicated the following research-findings: (1) WBB plays a role in facilitating positive participation in classes. (2) In contrast to the scenario of the traditional mode of instruction (without the usage of WBB), students were able to deepen their understanding of the theme by accessing the WBB before and after classes. (3) Peer evaluation highly promoted students' motivation to learn, and was effective in cultivating meta-cognition through modeling. (4) Mobile phone was identified as a highly effective tool for keeping class attendance, realizing interactive classes by generating discussions, and managing assignments and homework.

• Convergence Security Journal
• /
• v.24 no.2
• /
• pp.47-53
• /
• 2024

The traditional personal information life cycle model, primarily tailored to conventional systems, is inherently unsuitable for comprehending the nuances of personal information flow within artificial intelligence frameworks and for formulating effective protective measures. Therefore, this study endeavors to introduce a personal information life cycle model specifically designed for artificial intelligence (AI). This paper presents a personal information life cycle model suitable for artificial intelligence, which includes the stages of collection, retention, learning, use, and destruction/suspension, along with the re-learning process for destruction/suspension. Subsequently, we compare the performance of these existing models (such aspersonal information impact assessment and the ISMS-P model) with the newly proposed model. This underscores the superiority of our proposed model in comprehensively understanding the personal information flow in AI and establishing robust protective measures.

• Journal of Animal Science and Technology
• /
• v.66 no.1
• /
• pp.167-177
• /
• 2024

Iris segmentation is an initial step for identifying the biometrics of animals when establishing a traceability system for livestock. In this study, we propose a deep learning framework for pixel-wise segmentation of bovine iris with a minimized use of annotation labels utilizing the BovineAAEyes80 public dataset. The proposed image segmentation framework encompasses data collection, data preparation, data augmentation selection, training of 15 deep neural network (DNN) models with varying encoder backbones and segmentation decoder DNNs, and evaluation of the models using multiple metrics and graphical segmentation results. This framework aims to provide comprehensive and in-depth information on each model's training and testing outcomes to optimize bovine iris segmentation performance. In the experiment, U-Net with a VGG16 backbone was identified as the optimal combination of encoder and decoder models for the dataset, achieving an accuracy and dice coefficient score of 99.50% and 98.35%, respectively. Notably, the selected model accurately segmented even corrupted images without proper annotation data. This study contributes to the advancement of iris segmentation and the establishment of a reliable DNN training framework.

• Journal of The Korean Association For Science Education
• /
• v.18 no.2
• /
• pp.173-182
• /
• 1998

This study investigated the influences of metacognitive learning strategies upon 6th-graders' achievement, science process skill, use of cognitive strategies, use of metacognitive strategies, self-efficacy, intrinsic value, attitude toward science class, and scientific attitude. The metacognitive learning strategies were developed on the basis of previous results and modified in a pilot study. Before the instructions, a pretest of motivation was administered, and used as a blocking variable. The score of previous achievement test was used as covariates for achievement and science process skill. Tests of use of cognitive strategies, use of metacognitive strategies, self-efficacy, intrinsic value, attitude toward science class, and scientific attitude were also administered, and their scores were used as covariates. After the instructions, a researcher-made achievement test, the Middle Grades Integrated Science Process Skills Test, and post-tests of above variables were administrated. Two-way ANCOVA results revealed that the scores of the treatment group were significantly higher than those of the control group for all tests except for science process skill. No interactions between the treatment and the level of the previous motivation were found. Educational implications are discussed.

• Journal of The Korean Association of Information Education
• /
• v.18 no.1
• /
• pp.75-87
• /
• 2014

The purpose of this study was to investigate the educational usefulness, problems of smartpad-based instruction in elementary school based on qualitative research. To accomplish the purpose of the study, D elementary school in metropolitan area, two classes of digital textbook model school for seven years, were selected as the classes for observation. Six fourth and fifth graders and their teachers were interviewed and their 10 lessons were used for analyzing teaching and learning activities in smartpad-based instruction. The results of the study, 'Facilitating collaboration and interaction', 'effective use of various resource and SNS', 'improving concentrativeness', 'shortening waiting time' were identified as main educational usefulness. 'Lack of learning supporting tools', 'focusing on function rather than learning contents', 'low learning effectiveness', 'interference of flow' were presented as problems in using smartpad in classroom. Finally, some practical tasks for effective application of smartpad-based instruction were recommended.

• Journal of The Korean Association of Information Education
• /
• v.20 no.5
• /
• pp.499-506
• /
• 2016

Multi-cultural students have a variety of problems, such as the lack of Korean communication skills, learning slump and psychological anxiety. In order to solve these problems, It is developed to design a software education program for learning attitude, friendship interaction and sociality. It is developed on the basis of the major steps in the ADDIE model, Use-Needs-Design-Implementation-Share for multi-cultural elementary school students. To analyze the effects of software education Program, we chose the 15 elementary school students of 4th, 5th and 6th grade and adapted the program. Then, we analyzed the educational effects through the results of pre to post tests. Consequently, the software education program developed for this research revealed that it affected the learning attitude, friendship, sociality and programming interest of multi-cultural students.

• Journal of the Korean Society for information Management
• /
• v.25 no.3
• /
• pp.27-39
• /
• 2008

In bibliographic data, the use of personal names to indicate authors makes it difficult to specify a particular author since there are numerous authors whose personal names are the same. Resolving same-name author instances into different individuals is called author resolution, which consists of two steps: calculating author similarities and then clustering same-name author instances into different person groups. Author similarities are computed from similarities of author-related bibliographic features such as coauthors, titles of papers, publication information, using supervised or unsupervised methods. Supervised approaches employ machine learning techniques to automatically learn the author similarity function from author-resolved training samples. So far however, a few machine learning methods have been investigated for author resolution. This paper provides a comparative evaluation of a variety of recent high-performing machine learning techniques on author disambiguation, and compares several methods of processing author disambiguation features such as coauthors and titles of papers.