Earticle

This study aims to clarify the relationship between subjective perception and performance in the communicative learning using Synchronous Computer-Mediated Communication (SCMC). In this study, we developed four types of SCMC: videoconferencing (image and voice), audioconferencing (no image and voice), text-chat with image (image and no voice) and plain text-chat (no image and no voice). In this experiment, each system aids learners in being aware of and uttering a target formulaic expression as learning objectives. We investigated the effect of each system on psychological perception and productive performance, as well as the relationship between perception and performance. The results show that image and voice promote consciousness of natural communication and relief, while a text-mediated system enhances confidence in grammatical accuracy. In order to clarify the relationship between perception and performance, multiple regression analysis was conducted using perception as the independent variable and performance as the dependent variable. The results indicate that consciousness of natural communication and the use of the voice communication affect factors such as the number of utterances, while interlocutor’s image and voice and consciousness of grammatical accuracy lead to self-correction.

코로나19뿐만 아니라 IT 산업의 빠른 변화와 성장에 맞물려 온라인 교육도 발전이 급격하게 진행되므로 온라인 교육의 역할은 강조되어야 한다. 이와 같은 변화는 오프라인 중심으로 운영되던 대학들은 온라인 강의에 대한 경험 부족으로 질 높은 강의를 제공하는 데 어려움을 겪었다. 따라서 온라인 교육에 있어서 학습자의 학습성과와 학습만족을 높일 수 있는 방안을 모색할 필요가 있다. 본 연구는 온라인 교육서비스 품질이 학습몰입을 통해 학습성과와 학습만족에 어떠한 영향을 미치는지 분석함으로서 교육성과를 이끌어 낼 수 있는 요인들을 밝혀내고자 하였다. 연구의 목적을 달성하기 위해 중국인 한국어 학습자들을 대상으 로 Smart-PLS 방법을 통해 온라인 교육서비스 품질이 학습몰입, 학습성과 및 학습만족간의 영향관계를 검 증하였다. 분석 결과는 다음과 같다. 첫째, 온라인 교육서비스 품질이 학습만족에 긍정적인 정(+)의 영향을 미치는 것을 검정하였으나 학습성 과에 부(-)의 영향을 미치는 것으로 나타났다. 둘째, 학습몰입이 학습성과 및 학습만족에 긍정적인 정(+)의 영향을 미치는 것을 검정하였다. 셋째, 학습몰입이 온라인 교육서비스 품질이 학습성과 및 학습만족에 미치 는 영향에 있어 매개효과가 있는 것을 검정하였다. 따라서 연구 결과를 바탕으로 온라인 교육서비스 품질 및 학습몰입이 높이기 위해 실증적으로 방안을 제시하였고 온라인 교육서비스 품질이 학습성과에 촉진하는 유효적인 방법을 탐색하였다.

The role of online education should be emphasized as the development of online education is rapidly progressing in line with the rapid change and growth of the IT industry as well as COVID-19. This change was difficult for universities and academies, which were operated offline, to provide high-quality lectures due to their lack of experience in online lectures. Therefore, it is necessary to seek ways to increase learners' learning outcomes and learning satisfaction in online education. This study attempted to identify factors that can lead to educational performance by analyzing how online educational service quality affects learning performance and learning satisfaction through learning commitment. In order to achieve the purpose of the study, the influence relationship between learning commitment, learning performance, and learning satisfaction was verified through the Smart-PLS method for Chinese Korean language learners. The analysis results are as follows. First, it was tested that online education service quality had a positive (+) effect on learning satisfaction, but it was found to have a negative (-) effect on learning performance. Second, it was tested that learning commitment had a positive (+) effect on learning performance and learning satisfaction. Third, it was tested that learning commitment has a mediating effect on the effect of online education service quality on learning performance and learning satisfaction. Therefore, based on the research results, a plan was empirically proposed to increase the quality of online education service and learning commitment, and an effective method for promoting online education service quality in learning outcomes was explored.

The purpose of this study is to examine the effect of psychological capital on empowerment and learning performance and the mediating effect of the empowerment on the relationship between psychological capital and learning performance. In order to verify the relationships and mediating effect, data were obtained from 283 university students in Ulsan Metropolitan City and were analyzed by using SPSS 12.0, AMOS 5.0. The findings are as follows: First, the psychological capital were positively related to the empowerment and the learning performance. Second, there was also a positive relationship between the empowerment and the learning performance. Finally, it is found that empowerment fully mediated the relationship of psychological capital and earning performance. The theoretical implication of the study includes that this study and findings advance the understanding of learning performance by suggesting a new viewpoint regarding how psychological capital and empowerment to motivate university's learning performance. Based on these findings, the implications and the limitations of the study were presented including some directions for future studies.

인공지능 기술의 발달로 인해 사물인터넷 기기와 스마트 기기들의 사용이 급속도로 증가하였다. 기존의 인공지능 모델 학습 방법은 보안 문제 등 많은 단점이 존재한다. 따라서 기존의 단점을 보완하면서 원 데이터 송 신이 필요 없는 연합학습(FL: Federated Learning)의 연구가 활발히 이루어지고 있다. 연합학습은 많은 장점들을 가지고 있지만 참여 기기 간 통신 및 네트워크의 상황에 차이가 존재하기 때문에 일부 기기로 인해 전체적인 학 습이 지연될 수 있다는 한계를 가진다. 따라서 본 논문에서는 통신과 네트워크 상황에 따라 적절하게 학습 참여 기기를 선택하는 기법을 제안한다. 제안한 기법을 활용할 경우 측정된 네트워크와 통신 성능 지표를 바탕으로 각 기기들의 통신 및 네트워크 상황을 예측하고 이를 통해 기기들을 선택함으로써 보다 효율적인 연합학습을 수행한 다. 이는 기존의 학습기법과 비교하였을 때 전체적인 학습시간을 감소시키는 결과를 보였으며 더욱 짧은 시간에 더 높은 성능을 가지는 모델을 구축하는 결과를 보였다.

Supported by the development of artificial intelligence (AI) technology, the use of Internet of Things (IoT) and smart devices has increased rapidly. Existing methods of training AI models have many disadvantages, such as security problems. Therefore, many researches on federated learning (FL) that does not require transmission of raw data have been actively conducted for overcoming the existing shortcomings. FL has many advantages, but there is a limit that some devices may delay the FL system’s overall training due to the differences in communication and network situations between participating devices. To overcome such limitation, in this paper, we propose the participant selection scheme based on network situation prediction of server-to-device communication paths. Using the proposed scheme, more efficient FL can be performed by selecting participant devices intelligently based on the predicted communication and network information of each device.

기술의 발전으로 사물인터넷은 스스로 판단하고 동작을 수행하는 인공지능 기반의 지능형 사물인터넷으 로 발전하고 있다. 연합학습은 지능형 사물인터넷에 적합한 학습 방법으로서, 데이터가 탈중앙화된 상황에서 다수 의 사물인터넷 기기들과 서버가 협력하여 글로벌 모델을 학습시킨다. 연합학습은 학습을 수행하기 위해 통신 과정 이 필수적으로 포함되므로 효율적인 통신은 곧 연합학습의 성능향상으로 직결된다. 본 논문에서는 통신상황에 따 른 연합학습에서의 모델 전송 및 학습 라운드 진행에 대한 분석을 수행하였다. 그 과정에서 연합학습에서의 통신 상황에 따른 참여 기기들의 전송 소요 시간, 전송 로컬 모델 수 및 글로벌 모델 전송에 대한 분석을 수행하였다. 그 결과, 연합학습을 수행하기 위한 여러 과정들에서 많은 양의 통신이 반복적으로 수행되기 때문에 통신상황은 연합학습 성능에 영향을 주는 주요한 요소임을 보였으며, 학습에 참여하는 최소 기기 수와 한 라운드 소요 시간의 상한이 통신상황에 따라 적절하게 조절되어야 지연 없는 연합학습 수행이 가능함을 보였다.

Federated learning (FL) is a learning method suitable for the intelligent Internet of Things (IoT), and in a situation where data is decentralized, a number of IoT devices and a server cooperate to build a global model in FL. Since FL includes communication processes to perform learning, efficient communication is connected directly to better performance of FL. Therefore, in this paper, we analyzed the transmission of learning models and the learning round progress of FL by considering the communication situation. We analyzed the transmission delay of devices and the transmissions of local and global models depending on the communication situation in FL. As a result, the analysis shows that the communication situation is the major factor affecting FL performance because large amounts of communication are repeated frequently in FL and that it is important to adjust various parameters of FL properly by considering the communication situation.

This study investigates energy efficiency issues of device-to-device (D2D) communications in heterogeneous networks. To minimize the total transmitted power, an approach based on Q-learning together with adaptive ? -greedy is proposed to optimize the connection of user equipment (UE) with base station (BS) or access point (AP). The proposed adaptive ? -greedy can conduct the adequate exploration and exploitation operations for effective optimization. Simulation results indicate that in the single-cell scenario, the proposed method can attain performance close to the best solution.

Background Frailty is an important condition to detect in its early stages to prevent progression to more severe stages in older adults. Age-related declines in physical performance are strongly associated with frailty. Purpose This study aims to develop a frailty classification model by comparing the performance of machine learning models based on physical performance measures in community-dwelling older adults. Study design A cross-sectional study Methods Physical performance data were collected from older adults aged ≥65 years. Frailty classification models were developed using logistic regression, support vector machine (SVM), K-nearest neighbors (KNN), decision tree, and random forest. Clinical features including short physical performance battery, single-leg stance, SARC-F, body mass index, and mini-mental state examination (MMSE) were used as input variables for model development. The performance of each model was evaluated using accuracy, sensitivity, specificity, precision, F1-score, and area under the receiver operating characteristic curve (AUC). Permutation feature importance was employed to identify key predictors of frailty. Results The KNN model demonstrated the highest classification performance, achieving an accuracy of 0.93, an F1-score of 0.95, and an AUC of 0.86, indicating its suitability for frailty assessment. The logistic regression model achieved an accuracy of 0.86, an F1-score of 0.89, and an AUC of 0.98. The random forest model showed similar results, with an accuracy of 0.86, an F1-score of 0.88, and an AUC of 0.96. The SVM model recorded an accuracy of 0.79, an F1-score of 0.84, and an AUC of 0.80. The decision tree model showed the lowest performance, with an accuracy of 0.71, an F1-score of 0.78, and an AUC of 0.64. Feature importance analysis revealed that MMSE and SARC-F were the most influential predictors in the KNN model. Conclusions This study demonstrates that KNN is well-suited for identifying subtle variations in physical function that contribute to frailty. The results highlight its potential for clinical implementation in automated frailty screening. Feature importance analysis provides insight into key predictors, supporting personalized assessment strategies. However, due to the small sample size, further research is needed to assess the generalizability of frailty classification models in larger populations.

Background Logistics service workers face elevated injury rates, with ankle-related incidents comprising 23% of workplace injuries. Traditional binary foot posture classifications may oversimplify the complex biomechanical relationships affecting dynamic balance performance in occupational settings. Purpose To apply unsupervised machine learning clustering to identify distinct foot posture phenotypes among logistics service workers and compare dynamic balance performance between identified clusters using the Y-Balance Test. Study design Cross-sectional observational study Methods A total of 112 logistics service workers were analyzed using K-means clustering based on age, body mass index, work duration, navicular drop, and resting calcaneal stance position. Silhouette score analysis determined optimal cluster number. Dynamic balance was assessed using the Y-Balance Test, measuring reach distances in anterior, posteromedial, and posterolateral directions. Results Four distinct phenotypes emerged: rearfoot valgus-dominant pronated (n=22), midfoot collapse-dominant pronated (n=32), age-related (n=23), and supinated (n=35) foot types. Significant differences in dynamic balance performance were observed in the posterolateral direction (F=3.900, p=0.011). The supinated phenotype demonstrated superior posterolateral reach performance (77.76±13.63%) compared to midfoot collapse-dominant (66.18±16.37%, p=0.003) and age-related phenotypes (67.78±14.97%, p=0.018). Conclusions Unsupervised machine learning successfully identified naturally occurring foot posture phenotypes with distinct dynamic balance characteristics. Midfoot collapse-dominant pronation demonstrated greater balance impairments than rearfoot valgus patterns, supporting the implementation of phenotype-specific interventions for workplace injury prevention in logistics workers.

This paper investigates the Contention Window (CW) optimization problem in multi-agent scenarios, where the fully cooperative among mobile stations is considered. A partially observable environment is employed to model and analyze the CW optimization problem, and Smart Exponential-Threshold-Linear with Deep Q-learning Network (SETL-DQN) Multi-Agent (MA) algorithm is proposed to obtain the optimal system throughput through the CW Threshold optimization. In the determined scenarios, SETL-DQN(MA) can effectively cope with the mutual interaction among mobile stations. The simulation results show that our proposed method is superior from both static and dynamic scenarios and has the highest optimum packet transmission efficiency.

Many organizations have had deep interests in studies concerning leadership and in academic areas, in not only management but also psychology. Until now, leadership has been accentuated by managers or team leaders especially. Recently, however, the concept of self-leadership directing one"s own activities through self-control or self-management is being focused on practices and in academia. This study is to investigate the influence between self-leadership strategies and learning performance in IT classes mediated by attitude of attendance focused on the social science students in a university. Research results can give us direction of task-taking attitudes in firms or learning attitudes in teaching organizations and implications to human resource managers who are in charge of improving learning performance or productivity.

In the recent e-learning environment, avatars are often used to help learners get familiar with the contents, which is ultimately to motivate them to study more. Therefore, it is important to investigate whether avatars have actually the desirable effect on users of e-learning materials. Surprisingly, however, no extensive study has been conducted on this crucial issue Accordingly, main objectives this study are summarized as follows. First, we need to gain better understanding of how much learners’ trust towards avatars (termed as “avatar trust”) is transferred to learners’ trust towards e-learning contents (termed as “contents trust”). Second, we need to investigate how much learners’ personal relationships with avatars as well as learning behaviors change depending on avatar types (attractive vs. professional) and contents complexity (easy vs. difficult). As described in the study objectives, in order for us to analyze empirical data more systematically, we classified avatar types into two: “attractive” and “professional;” the contents are categorized as either “easy” or “difficult.” Therefore, it is essential for this study to build a prototype e-learning website on which our research purpose can be realized and tested effectively with proper avatar types and e-learning contents. For this purpose, we built a prototype e-learning website, in which avatars are invited from currently working avatar instructors used in real-world e-learning websites, and e-learning contents are adapted from real-world contents about Java programming topic, which have been proved to have shown high quality and reliability. Our research method includes questionnaire survey by inviting a number of valid respondents comprised of office workers who are believed to have high demands for the e-learning contents as well as those who have previous experience with avatar instructors. Respondents were given one of the four e-learning experiment conditions (2 avatar types x 2 contents types) on a random basis. Each experimental e-learning condition is framed to have the same quality but different avatar type and content complexity. Then the respondents are asked to fill out the survey form which has questions about avatar trust, contents trust, personal relationships with avatar, and learning behavior, among others. Regarding the constructs used in research model, we based them rigorously on previous studies. For example, we used six constructs such as behavior to give information (BGI), behavior to obtain information (BOI), need for inclusion wanted, need for control wanted, contents trust, and avatar trust. To measure them, 7-Likert scales were used in the questionnaire. E-learning performance was measured indirectly through two constructs such as BGI and BOI. Six constructs used in the research model were adopted and revised from the FIRO-B model suggested by Schutz. Empirical results are as follows: First, professional avatars are more effective for difficult contents, while attractive avatars were not as effective for easy contents. Second, our study results ascertained that avatar trust transfers to contents trust regardless of avatar types and contents complexity.

According to a lot of people, home literacy practices are closely related to language learners’ school performance. Therefore, for more successful language learning the relationships between both of them are required to investigate. In this study, particularly the relationships between home literacy practices and school performance in bilingual and multilingual contexts have demonstrated. A variety of literacy practices involving family literacy practices, community cultural literacy, and social background were discussed. A number of studies showed that family literacy practices in first and second languages influence language learners’ literacy learning and school performance positively or negatively. Thus, the role of families and parents at home for language-minority learners’ literacy development and academic success in school might be very important.

This study explores the effect of web-based collaborative writing activities through interactive peer-to-peer corrective feedback conducted by students. 40 university students participated in a writing module and were encouraged to participate in both individual and group writing activities based on online interactions with other peers. 40 writing samples from individual writing tasks and 40 writing samples from the web-based writing were collected and scored by both a native teacher and a Korean teacher. Cohen’s analytic scoring rubric (1994) was used for the analysis of learners’ corrective behaviors. This study illustrates three main research questions on student writing behaviors through web-based collaborative Business writing activities. This study shows that experiencing online collaborative Business writing activities may have positive effects on both students’ writing and their writing strategies.

인터넷 서비스의 확산과 함께 정교해지는 피싱(Phishing) 공격은 개인 정보 탈취 및 금융 피해를 유발하는 심각한 보안 위 협으로 대두되고 있다. 기존의 피싱 탐지 체계는 주로 구글 세이프 브라우징(Google Safe Browsing)이나 피쉬탱크 (PhishTank)와 같은 블랙리스트(Blacklist) 방식에 의존해 왔다. 이 방식은 알려진 위협에 대해서는 신속하고 정확한 차단이 가능하나, 제로데이(Zero-day) 공격을 탐지하지 못하는 치명적인 한계를 가진다. 본 연구에서는 이러한 한계를 극복하기 위해 URL의 어휘적 특징을 기반으로 하는 다양한 인공지능 모델의 탐지 성능을 비교 분석하였다. 실험 대상 모델로는 전통적인 휴 리스틱 알고리즘과 머신러닝 모델인 로지스틱 회귀(Logistic Regression), 서포트 벡터 머신(SVM), 랜덤 포레스트(Random Forest), 그리고 딥러닝 모델인 CNN(1D)과 LSTM을 선정하였다. 실험 결과, 휴리스틱 방식은 44.5%의 저조한 정확도를 보인 반면, SVM(RBF 커널) 모델은 97.0%의 정확도와 0.970의 F1-Score를 기록하며 가장 우수한 성능을 나타냈다. 특히 딥러닝 모 델인 CNN(94.5%)과 LSTM(76.1%) 대비 SVM은 0.165초라는 빠른 추론 속도를 보여 실시간 탐지 환경에서 성능과 효율성의 최적 균형을 갖춘 모델임을 입증하였다.

As internet services proliferate, phishing attacks are becoming increasingly sophisticated and are emerging as a serious security threat that causes the theft of personal information and financial damage. Existing phishing detection systems have primarily relied on blacklist methods such as Google Safe Browsing or PhishTank. While this approach enables the rapid and accurate blocking of known threats, it has a critical limitation in its inability to detect zero-day attacks. To overcome these limitations, this study comparatively analyzed the detection performance of various artificial intelligence models based on the lexical features of URLs. The models selected for the experiment included traditional heuristic algorithms, machine learning models such as Logistic Regression, Support Vector Machine(SVM), and Random Forest, as well as deep learning models like CNN(1D) and LSTM. The experimental results showed that while the heuristic method yielded a poor accuracy of 44.5%, the SVM(RBF kernel) model demonstrated the superior performance, recording an accuracy of 97.0% and an F1-Score of 0.970. In particular, compared to the deep learning models CNN(94.5%) and LSTM(76.4%), SVM demonstrated a fast inference speed of 0.165 seconds, proving it to be the model with the optimal balance between performance and efficiency in a real-time detection environment.

The purpose of this study is to investigate the effects of mobile-learning (m-learning) on EFL student satisfaction and performance. For this empirical study, a research model was proposed with hypotheses. For m-learning attributes, four variables were suggested: Mobility, Content Personalization, Interactivity, and Accessibility. In addition, Learning Supports mechanisms used as potential moderating factor was proposed in order to test whether this variable enhances the relationship between Satisfaction and Performance in English m-learning. the data were collected from actual m-learning EFL students. A total of 279 responses were analyzed to test the proposed hypotheses. Results show that all hypotheses were significantly supported, suggesting that m-learning attributes are crucial factors for increasing EFL learners’ satisfaction and performance. Furthermore, the relationship between Satisfaction and Performance is enhanced by the inclusion of various types of Learning Supports. Finally, academic and practical implications are discussed.

This study empirically verified the factors affecting the learning outcomes of English as a Foreign Language (EFL) learners in a smart learning (s-learning) environment. Based on social cognitive theory (SCT), six independent constructs (facilitating conditions, social influence, positive attitude toward s-learning, self-efficacy, social interaction, and knowledge) were proposed in the research model. The data collected from 455 s-learning EFL learners were analyzed in order to test the causal relationships between these six constructs and s-learning performance. For the analysis, the structural equation model (SEM) with AMOS 22.0 was used, and all six constructs were found to have a significant effect on s-learning performance. Important implications are discussed based on the study results.

This paper discusses the performance of linear regression, regression tree, support vector regression, and ensemble learners in modelling airflow between two spaces based on accuracy and training time. To obtain training data, different scenarios from an existing computational fluid dynamics (CFD) model are simulated via transient analysis using Cradle scSTREAM. The raw dataset is transformed to having time step sizes of 2.5s, 5.0s, and 50.0s. Feature scaling is also employed on the each data set using both min-max scaling and z-score methods for a total of 9 datasets. Hyperparameters according to machine learning (ML) algorithms are varied such that 15 ML models across the four algorithms are developed. The results show that the regression trees perform the best over all other algorithms, with all models maintaining R2 values above 0.95 at the different datasets. On the other hand, as expected, all linear models demonstrated poor performance compared to nonlinear models. Data resolution affects model accuracy and training time, with accuracy declining slightly as time step size increased. It is also found that there is no significant effect of feature scaling. Lastly, ML models yield substantially cheaper simulation costs than CFD to simulate airflow.

실패 학습 행동은 창업 벤처의 성과를 향상시키는 중요한 방법이며, 본 연구는 실패 학습 행동이 창업 성과에 미치는 영향의 근본적인 메커니 즘을 탐구하는 데 전념합니다. 관련 이론을 참고하여 본 연구에서는 비즈니스 모델 혁신을 매개변수로, 조직 관성을 조절변수로 선택했습니다. 본 연구는 설문지를 통해 307개 스타트업 표본으로부터 데이터를 수집하고 층화 회귀분석을 통해 가설검증을 실시했습니다. 그 결과 실패 학습 행동은 재무적 및 비재무적 기업 성과에 유의미한 정(+)의 영향을 미치며, 비즈니스 모델 혁신은 이 효과를 부분적으로 매개하는 역할을 하는 것 으로 나타났습니다. 조직의 관성은 비즈니스 모델 혁신이 재무적 기업가적 성과에 미치는 영향을 약화시킵니다. 본 연구는 실패 학습 행동과 기 업가적 성과에 대한 이론적 결과를 확장하여 기업 경영 실무에 시사점을 제공합니다.

Failure learning behavior is an important way to promote the performance of entrepreneurial firms, and this study is dedicated to exploring the underlying mechanisms of the impact of failure learning behavior on entrepreneurial performance. Referring to related theories, this study selects business model innovation as a mediating variable and organizational inertia as a moderating variable. The study adopts a questionnaire method to collect data from a sample of 307 start-ups and conducts hypothesis testing through stratified regression. The results show that failure learning behavior has a significant positive effect on financial and non-financial entrepreneurial performance, and business model innovation assumes a partial mediating role in the effect. Organizational inertia negatively moderates the effect of business model innovation on financial entrepreneurial performance. This study extends the theoretical results on failure learning behavior and entrepreneurial performance and has implications for business management practice.

Background/Objective. The COVID-19 has accelerated the conduct of online classes around the world. To administer distant learning, schools are now relying on learning management systems. With the problems of the abrupt shift to an online modality, there must be measures in place to cluster students’ characteristics in order to assist their learning requirements, struggles, and academic performances. Methods/Statistical Analysis. Based on Delone and McLean’s IS Success model, this study used quantitative approaches and exploratory techniques to explain the higher education learners’ clusters. From March to April 2021, 303 samples were randomly selected from students participating in online programs at a higher education school on Taft Avenue in Manila. Sex, school, frequency of use, time duration, and experience with the LMS were all factors in the study. Both Hierarchical Cluster Analysis and K-Means Cluster Analysis were used to classify the samples. Findings. The results revealed that there are four clusters formed which are labeled as: service-oriented, system quality-oriented, holistic-oriented, and LMS-averse. Based on the results, information quality has the greatest influence. Since the clustering analysis is non-inferential and it is used as an exploratory technique, the researchers do not guarantee a unique solution as this depends on the elements of the subjects and the variables used. Improvements/Applications. The academic stakeholders can use the cluster analysis to make the appropriate interventions to improve the online course content and enhance the students’ academic performance. Some of these appropriate interventions include creating course policies, planning the necessary training of teachers and students, and developing the online delivery of the lessons.