• 한국과학교육학회지
• /
• 제42권2호
• /
• pp.177-184
• /
• 2022

This article aims to explain the concept and characteristics of the 'Intelligent Science Lab', which is being promoted nationwide in Korea since 2021. The Korean Ministry of Education creates a master plan containing a vision for science education every five years. The most recently announced '4th Master plan for science education (2020-2024)' emphasizes the policy of setting up an 'intelligent science lab' in all elementary and secondary schools as an online and offline space for scientific inquiry using advanced technologies, such as Internet of Things and Augmented and Virtual Reality. The 'Intelligent Science Lab' project is being pursued in two main directions: (1) developing an online platform named 'Intelligent Science Lab-ON' that supports science inquiry classes, and (2) building a science lab space in schools that encourages active student participation while utilizing the online platform. This article presents the key features of the 'Intelligent Science Lab-ON' and the characteristics of intelligent science lab spaces newly built in schools. Furthermore, it introduces inquiry-based science learning programs developed for intelligent science labs. These programs include scientific inquiry activities in which students generate and collect data ('data generation' type), utilize datasets provided by the online platform ('data utilization' type), or utilize open and public data sources ('open data source' type). The Intelligent Science Lab project is expected to not only encourage students to engage in scientific inquiry that solves individual and social problems based on real data, but also contribute to presenting a model of online and offline linked scientific inquiry lessons required in the post-COVID-19 era.

• 한국산업보건학회지
• /
• 제32권1호
• /
• pp.21-30
• /
• 2022

Objectives: The purpose of this study is to suggest an optimal method by comparing the analysis methods of work environment measurement datasets including left-censored data where one or more measurements are below the limit of detection (LOD). Methods: A computer program was used to generate left-censored datasets for various combinations of censoring rate (1% to 90%) and sample size (30 to 300). For the analysis of the censored data, the simple substitution method (LOD/2), β-substitution method, maximum likelihood estimation (MLE) method, Bayesian method, and regression on order statistics (ROS)were all compared. Each method was used to estimate four parameters of the log-normal distribution: (1) geometric mean (GM), (2) geometric standard deviation (GSD), (3) 95th percentile (X95), and (4) arithmetic mean (AM) for the censored dataset. The performance of each method was evaluated using relative bias and relative root mean squared error (rMSE). Results: In the case of the largest sample size (n=300), when the censoring rate was less than 40%, the relative bias and rMSE were small for all five methods. When the censoring rate was large (70%, 90%), the simple substitution method was inappropriate because the relative bias was the largest, regardless of the sample size. When the sample size was small and the censoring rate was large, the Bayesian method, the β-substitution method, and the MLE method showed the smallest relative bias. Conclusions: The accuracy and precision of all methods tended to increase as the sample size was larger and the censoring rate was smaller. The simple substitution method was inappropriate when the censoring rate was high, and the β-substitution method, MLE method, and Bayesian method can be widely applied.

• 대한임베디드공학회논문지
• /
• 제19권3호
• /
• pp.151-158
• /
• 2024

Although on-device artificial intelligence (AI) has gained attention to diagnosing machine faults in real time, most previous studies did not consider the model retraining and redeployment processes that must be performed in real-world industrial environments. Our study addresses this challenge by proposing an on-device AI-based real-time machine fault diagnosis system that utilizes continual learning. Our proposed system includes a lightweight convolutional neural network (CNN) model, a continual learning algorithm, and a real-time monitoring service. First, we developed a lightweight 1D CNN model to reduce the cost of model deployment and enable real-time inference on the target edge device with limited computing resources. We then compared the performance of five continual learning algorithms with three public bearing fault datasets and selected the most effective algorithm for our system. Finally, we implemented a real-time monitoring service using an open-source data visualization framework. In the performance comparison results between continual learning algorithms, we found that the replay-based algorithms outperformed the regularization-based algorithms, and the experience replay (ER) algorithm had the best diagnostic accuracy. We further tuned the number and length of data samples used for a memory buffer of the ER algorithm to maximize its performance. We confirmed that the performance of the ER algorithm becomes higher when a longer data length is used. Consequently, the proposed system showed an accuracy of 98.7%, while only 16.5% of the previous data was stored in memory buffer. Our lightweight CNN model was also able to diagnose a fault type of one data sample within 3.76 ms on the Raspberry Pi 4B device.

• 한국과학교육학회지
• /
• 제44권3호
• /
• pp.231-248
• /
• 2024

본 연구에서는 오픈소스 소프트웨어와 인공지능 문서 분류 모델인 한국어 Sentence-BERT로 고등학교 1학년 통합과학 질문-답변 챗봇을 제작하고 2023학년도 1년 동안 독립형 서버에서 운영했다. 챗봇은 Sentence-BERT 모델로 학생의 질문과 가장 유사한 질문-답변 쌍 6개를 찾아 캐러셀 형태로 출력한다. 질문-답변 데이터셋은 인터넷에 공개된 자료를 수집하여 초기 버전을 구축하였고, 챗봇을 1년 동안 운영하면서 학생의 의견과 사용성을 고려하여 자료를 정제하고 새로운 질문-답변 쌍을 추가했다. 2023학년도 말에는 총 30,819개의 데이터셋을 챗봇에 통합하였다. 학생은 챗봇을 1년 동안 총 3,457건 이용했다. 챗봇 사용 기록을 빈도분석 및 시계열 분석한 결과 학생은 수업 중 교사가 챗봇 사용을 유도할 때 챗봇을 이용했고 평소에는 방과 후에 자습하면서 챗봇을 활용했다. 학생은 챗봇에 한 번 접속하여 평균적으로 2.1~2.2회 정도 질문했고, 주로 사용한 기기는 휴대폰이었다. 학생이 챗봇에 입력한 용어를 추출하고자 한국어 형태소 분석기로 명사와 용언을 추출하여 텍스트 마이닝을 진행한 결과 학생은 과학 질문 외에도 시험 범위 등의 학교생활과 관련된 용어를 자주 입력했다. 학생이 챗봇에 자주 물어본 주제를 추출하고자 Sentence-BERT 기반의 BERTopic으로 학생의 질문을 두 차례 범주화하여 토픽 모델링을 진행했다. 전체 질문 중 88%가 35가지 주제로 수렴되었고, 학생이 챗봇에 주로 물어보는 주제를 추출할 수 있었다. 학년말에 학생을 대상으로 한 설문에서 챗봇이 캐러셀 형태로 결과를 출력하는 형태가 학습에 효과적이었고, 통합과학 학습과 학습 목적 이외의 궁금증이나 학교생활과 관련된 물음에 답해주는 역할을 수행했음을 확인할 수 있었다. 본 연구는 공교육 현장에서 학생이 실제로 활용하기에 적합한 챗봇을 개발하여 학생이 장기간에 걸쳐 챗봇을 사용하는 과정에서 얻은 데이터를 분석함으로써 학생의 요구를 충족할 수 있는 챗봇의 교육적 활용 가능성을 확인했다는 점에 의의가 있다.