• Journal of Engineering Education Research
• /
• v.24 no.3
• /
• pp.21-31
• /
• 2021

Flipped learning, which involves listening to lectures at home and performing dynamic group-based problem-solving activities in the classroom, is recently evaluated as a learner-centered teaching method, and interest and applications in engineering education are increasing. Therefore, this study aims to provide practical guidelines for successful application through empirical research analysis on the use of flipped learning in domestic engineering education. Through the selection criteria and keyword search, a systematic review of 36 articles was conducted. As a result of the analysis, flipped learning research in engineering education has increased sharply since 2016, focusing on academic journals and reporting its application cases and effects. Most of the research supported that flipped learning was effective not only for learners' learning activities(e.g., academic achievement, satisfaction, engagement, learning-flow, interaction), but also for individualized learning and securing sufficient practice time. It was often used in major classes with 15 to less than 50 students, especially in computer-related major courses. Most of them consisted of watching lecture videos, active learning activities, and lectures by instructors, and showed differences in management strategies for each class type. Based on the analysis results, suggestions for effective flipped learning management in future engineering education were presented.

• Annual Conference on Human and Language Technology
• /
• 2021.10a
• /
• pp.55-60
• /
• 2021

최근 한국어에 대한 자연어 처리 연구는 딥러닝 기반의 자연어 이해 모델을 중심으로 각 모델의 성능에 대한 비교 분석과 평가가 활발하게 이루어지고 있다. 그러나 한국어 생성 모델에 대해서도 자연어 이해 영역의 하위 과제(e.g. 감정 분류, 문장 유사도 측정 등)에 대한 수행 능력만을 정량적으로 평가하여, 생성 모델의 한국어 문장 구성 능력이나 상식 추론 과정을 충분히 평가하지 못하고 있다. 또한 대부분의 생성 모델은 여전히 간단하고 일반적인 상식에 부합하는 자연스러운 문장을 생성하는 것에도 큰 어려움을 겪고 있기에 이를 해결하기 위한 개선 연구가 필요한 상황이다. 따라서 본 논문은 이러한 문제를 해결하기 위해 한국어 생성 모델이 일반 상식 추론 능력을 바탕으로 문장을 생성하도록 KommonGen 데이터셋을 제안한다. 그리고 KommonGen을 통해 한국어 생성 모델의 성능을 정량적으로 비교 분석할 수 있도록 평가 기준을 구성하고, 한국어 기반 자연어 생성 모델의 개선 방향을 제시하고자 한다.

• Journal of Intelligence and Information Systems
• /
• v.29 no.1
• /
• pp.249-263
• /
• 2023

Collaborative Filtering, a representative recommendation system methodology, consists of two approaches: neighbor methods and latent factor models. Among these, the latent factor model using matrix factorization decomposes the user-item interaction matrix into two lower-dimensional rectangular matrices, predicting the item's rating through the product of these matrices. Due to the factor vectors inferred from rating patterns capturing user and item characteristics, this method is superior in scalability, accuracy, and flexibility compared to neighbor-based methods. However, it has a fundamental drawback: the need to reflect the diversity of preferences of different individuals for items with no ratings. This limitation leads to repetitive and inaccurate recommendations. The Adaptive Deep Latent Factor Model (ADLFM) was developed to address this issue. This model adaptively learns the preferences for each item by using the item description, which provides a detailed summary and explanation of the item. ADLFM takes in item description as input, calculates latent vectors of the user and item, and presents a method that can reflect personal diversity using an attention score. However, due to the requirement of a dataset that includes item descriptions, the domain that can apply ADLFM is limited, resulting in generalization limitations. This study proposes a Generalized Adaptive Deep Latent Factor Recommendation Model, G-ADLFRM, to improve the limitations of ADLFM. Firstly, we use item ID, commonly used in recommendation systems, as input instead of the item description. Additionally, we apply improved deep learning model structures such as Self-Attention, Multi-head Attention, and Multi-Conv1D. We conducted experiments on various datasets with input and model structure changes. The results showed that when only the input was changed, MAE increased slightly compared to ADLFM due to accompanying information loss, resulting in decreased recommendation performance. However, the average learning speed per epoch significantly improved as the amount of information to be processed decreased. When both the input and the model structure were changed, the best-performing Multi-Conv1d structure showed similar performance to ADLFM, sufficiently counteracting the information loss caused by the input change. We conclude that G-ADLFRM is a new, lightweight, and generalizable model that maintains the performance of the existing ADLFM while enabling fast learning and inference.

• Journal of the Korea Convergence Society
• /
• v.11 no.6
• /
• pp.51-57
• /
• 2020

Children with language developmental disabilities often struggle through their lives from a lot of challenges in everyday life and social activities. They're often easily deprived of the opportunity to engage in social activities, because they find difficulty in understanding or using language, a core means of communication. With regard to this issue, AAC(Augmentative and Alternative Communication) can be an effective communication tool for children who are suffering from language disabilities. In this paper, we propose a deep learning-based AI service to make full use of the pictogram as an AAC tool for children with language developmental disabilities to improve not only the ability to interact with others but the capacity to understand language. Using this service, we strive to help these children to more effectively communicate their intention or desire and enhance the quality of life.

• Journal of The Korean Association of Information Education
• /
• v.11 no.2
• /
• pp.195-203
• /
• 2007

Together with the development of ubiquitous computing technology and robot technology, intelligent type robots are being utilized in many areas and the range is expected to become wider and wider. Among many service robots, the educational robots are being diversely studied in the field with the concept of r-Learning. Presently, teaching assistant robots require a lot of HRI studies prior to their practical use in the very near future and are not sufficient yet. Especially, the reactions of students based of the styles of robots (e.g. serious robot, playful robot) are very important in producing robot contents but there has been no case study on them. Therefore, in this study, the appearance of the IROBI was changed to become a teaching assistant robot and was used to test and compare elementary school students' interest, achievement and concentration depending on the styles of robot (playful, serious). The results showed that the interest was high in the group that had studied with the playful robot. The achievement however, did not show significant relations with the style of robot and that the concentration was high in the group that had studied with the serious robot.

• The Journal of Korean Academic Society of Nursing Education
• /
• v.21 no.1
• /
• pp.119-128
• /
• 2015

Purpose: This study was conducted to evaluate the effectiveness of an evidence-based nursing (EBN) course using action learning-based team learning in undergraduate nursing students. Methods: A quasi-experimental pretest-posttest control group design was employed. The participants who consented were 45 second-year nursing students (22 in the experimental, 23 in the control group) from a university in G-city, Korea. The intervention included lectures, practicals, team activities and reflection on overviewing EBN, formulating clinical questions, searching the evidence, and criticizing the research articles. At the beginning and the end of the 7-week EBN course, the participants completed self-reported questionnaires. Frequencies, $x^2$-test, t-test, and ANCOVA with the SPSS program 18.0, were used to analyze the data. Results: The experimental group showed significantly higher scores on EBN competency (F=25.80, p<001), information literacy (F=13.75, p=.001), and proactivity in problem solving (F=5.32, p=.026) than the control group. Conclusion: This study provides evidence that an EBN course improves undergraduate nursing students' EBN competencies, information literacy, and proactivity in problem solving. Team learning in EBN education can be an effective teaching strategy.

• Journal of Digital Contents Society
• /
• v.19 no.5
• /
• pp.845-852
• /
• 2018

This study has developed a preventive launcher from smartphone addiction for the digital generation and the contents recommendation based on machine learning which used multiple and collective intelligence. This could provide convenient digital nurturing experience for the parents who fear their children's over use of digital devices and also suggest individually adaptive digital learning methods that enhance the learning efficiency and pleasurable and safe learning environment for the children. Suggested application is a kind of gamification launcher that protects children from harmful contents and from smartphone addiction with time limit settings. For parents who find difficulty choosing from various kinds of contents and applications for education, this suggested system could provide a learning analytic report based on big data after collecting and analyzing the data of their children's learning and activities and recommend contents necessary for their kids using recommended algorithm by collective intelligence.

• Journal of the Korea Society for Simulation
• /
• v.29 no.4
• /
• pp.1-7
• /
• 2020

Early detection and classification of crop diseases play significant role to help farmers to reduce disease spread and to increase agricultural productivity. Recently, many researchers have used deep learning techniques like convolutional neural network (CNN) classifier for crop disease inspection with dataset of crop leaf images (e.g., PlantVillage dataset). These researches present over 90% of classification accuracy for crop diseases, but they have ability to detect only the pre-trained diseases. This paper proposes an efficient disease inspection CNN model for new crops not used in the pre-trained model. First, we present a benchmark crop disease classifier (CDC) for the crops in PlantVillage dataset using VGG16. Then we build a modified crop disease classifier (mCDC) to inspect diseases for untrained crops. The performance evaluation results show that the proposed model outperforms the benchmark classifier.

• Fashion & Textile Research Journal
• /
• v.23 no.3
• /
• pp.390-405
• /
• 2021

Gait analysis trackers have been developed to monitor daily gait patterns to improve users' running performance and reduce the risk of injuries. A variety of gait analysis trackers are available on the market(e.g., foot pods, insoles). Depending on the type of gait analysis tracker, users' discomfort or satisfaction as well as required properties may differ. Hence, the purpose of this study was to compare and analyze user experience of three different types of commercial shoe-mounted gait analysis trackers and their mobile applications in a laboratory environment using questionnaires based on actual experiences of each product. Ten males and ten females who regularly enjoy walking and running exercises participated in the experiment. After the participants set up the tracker and application themselves without support from researchers, ten to thirty minutes' exercise was permitted on each product. Following this, the participants answered questionnaires containing evaluation variables on the device and mobile application, as well as satisfaction, intention to use, recommendation, and purchase. In addition, they were asked questions about the attractive features and shortcomings of each device and application. The results showed that the PRO-SPECS^® smart insole was preferred over the others for ease of use, perceived durability, psychological burden of the design, and usefulness of the information provided by the application. Along with the results of questionnaire, this study also discussed strategies and recommendations for future product design and development.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.46 no.1
• /
• pp.23-31
• /
• 2023

Research and interest in sustainable printing are increasing in the packaging printing industry. Currently, predicting the amount of ink required for each work is based on the experience and intuition of field workers. Suppose the amount of ink produced is more than necessary. In this case, the rest of the ink cannot be reused and is discarded, adversely affecting the company's productivity and environment. Nowadays, machine learning models can be used to figure out this problem. This study compares the ink usage prediction machine learning models. A simple linear regression model, Multiple Regression Analysis, cannot reflect the nonlinear relationship between the variables required for packaging printing, so there is a limit to accurately predicting the amount of ink needed. This study has established various prediction models which are based on CART (Classification and Regression Tree), such as Decision Tree, Random Forest, Gradient Boosting Machine, and XGBoost. The accuracy of the models is determined by the K-fold cross-validation. Error metrics such as root mean squared error, mean absolute error, and R-squared are employed to evaluate estimation models' correctness. Among these models, XGBoost model has the highest prediction accuracy and can reduce 2134 (g) of wasted ink for each work. Thus, this study motivates machine learning's potential to help advance productivity and protect the environment.