• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.15 no.3
• /
• pp.1100-1118
• /
• 2021

In a wide range of ML applications, the training data contains privacy-sensitive information that should be kept secure. Training the ML systems by privacy-sensitive data makes the ML model inherent to the data. As the structure of the model has been fine-tuned by training data, the model can be abused for accessing the data by the estimation in a reverse process called model inversion attack (MIA). Although, MIA has been applied to shallow neural network models of recognizers in literature and its threat in privacy violation has been approved, in the case of a deep learning (DL) model, its efficiency was under question. It was due to the complexity of a DL model structure, big number of DL model parameters, the huge size of training data, big number of registered users to a DL model and thereof big number of class labels. This research work first analyses the possibility of MIA on a deep learning model of a recognition system, namely a face recognizer. Second, despite the conventional MIA under the white box scenario of having partial access to the users' non-sensitive information in addition to the model structure, the MIA is implemented on a deep face recognition system by just having the model structure and parameters but not any user information. In this aspect, it is under a semi-white box scenario or in other words a gray-box scenario. The experimental results in targeting five registered users of a CNN-based face recognition system approve the possibility of regeneration of users' face images even for a deep model by MIA under a gray box scenario. Although, for some images the evaluation recognition score is low and the generated images are not easily recognizable, but for some other images the score is high and facial features of the targeted identities are observable. The objective and subjective evaluations demonstrate that privacy cyber-attack by MIA on a deep recognition system not only is feasible but also is a serious threat with increasing alert state in the future as there is considerable potential for integration more advanced ML techniques to MIA.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.19 no.12
• /
• pp.197-211
• /
• 2018

This is an exploratory study on the status of the social work field practicum at a cyber university. The purpose is to investigate the current situation and improvement plan of the social work field practicum. A qualitative analysis was conducted with 11 professors who have instructed the social work field practicum at cyber universities. The social work field practicum based on the experiences of the professors is investigated, and this paper analyzes the status according to students, schools, practitioners, and institutions. In order to improve the quality of the social work field practicum, factors for improvement were analyzed through the efforts of students, schools, the Korean social workers' association, institutional improvements, and social welfare instructors. The results of the study are as follows. Students, schools, and training organizations should recognize the importance of the social work field practicum and must strive for systematic and consistent education. It is also important to remember that a social worker is not a professional with simple qualifications, but an expert with a philosophy, values, and ideologies. The direction for improvement in the social work field practicum is as follows. When constructing a social welfare curriculum, the school should have a realistic curriculum and teaching method that can enhance the sense of the field. The student should not be qualified as a social worker only as a vague investment for the future, but should have the professional ability to serve clients as a social worker and to give professional help to clients, considering the best welfare service for human beings. Institutions should provide a place for students to integrate theory and practice in vital social welfare experiences as social workers. The Republic of Korea is now facing an age with one million social workers. In order to open the future of social welfare in Korea, we need united endeavors with government that can develop students as pre-social workers and establish universities, institutions, and their systems for a substantial social work field practicum.

• Korean Journal of Child Studies
• /
• v.30 no.6
• /
• pp.223-235
• /
• 2009

Elementary education in Korea has variously changed and developed during the last thirty years. Among all the progressive changes, the improvement of teaching conditions is considered to be the most fundamental one. The number of students assigned to a class or a teacher has decreased to a considerable extent. Cyber teaching-learning has been peformed at school and home, and English education has been emerged as a significant part of the Korean public education. The research issues constantly considered essential over the past three decades starting in the 1980's are those related to curricula, teaching-learning methods, training of teachers in-service, and education for upright characters. The practical and political issues should be dealt with to revise the three integrated subjects and text books into a sole integrated subject and text book, to double the credential of teachers in terms of professionalism and to decrease the number of students per teacher in Korea to the OECD level.

• Journal of Korea Multimedia Society
• /
• v.21 no.8
• /
• pp.952-959
• /
• 2018

Clustering is one of the most fundamental algorithms in machine learning. The performance of clustering is affected by the distribution of data, and when there are more data or more dimensions, the performance is degraded. For this reason, we use a stacked auto encoder, one of the deep learning algorithms, to reduce the dimension of data which generate a feature vector that best represents the input data. We use k-means, which is a famous algorithm, as a clustering. Sine the feature vector which reduced dimensions are also multi dimensional, we use the Euclidean distance as well as the cosine similarity to increase the performance which calculating the similarity between the center of the cluster and the data as a vector. A deep clustering networks combining a stacked auto encoder and k-means re-trains the networks when the k-means result changes. When re-training the networks, the loss function of the stacked auto encoder and the loss function of the k-means are combined to improve the performance and the stability of the network. Experiments of benchmark image ad document dataset empirically validated the power of the proposed algorithm.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.12 no.10
• /
• pp.5143-5158
• /
• 2018

Improving the intrusion detection system (IDS) is a pressing need for cyber security world. With the growth of computer networks, there are constantly daily new attacks. Machine Learning (ML) is one of the most important fields which have great contribution to address the intrusion detection issues. One of these issues relates to the imbalance of the diverse classes of network traffic. Accuracy paradox is a result of training ML algorithm with imbalanced classes. Most of the previous efforts concern improving the overall accuracy of these models which is truly important. However, even they improved the total accuracy of the system; it fell in the accuracy paradox. The seriousness of the threat caused by the minor classes and the pitfalls of the previous efforts to address this issue is the motive for this work. In this paper, we consolidated stratified sampling, cost function and weighted Support Vector Machine (WSVM) method to address the accuracy paradox of ID problem. This model achieved good results of total accuracy and superior results in the small classes like the User-To-Remote and Remote-To-Local attacks using the improved version of the benchmark dataset KDDCup99 which is called NSL-KDD.

• Journal of the Korea Society of Computer and Information
• /
• v.16 no.4
• /
• pp.197-204
• /
• 2011

In this research, we have developed a virtual interview simulation system that utilizes the 3D stereoscopic technology. For this virtual simulation can play individual question stereoscopic movies using seamless loop technology, it provides realistic environment and interviewee with 3D filmed interviewer increasing reliable experience for interviewees. Implementing question-pool system and real-time construction of questionnaires is also available so that the interviewee can train and prepare for the various situations. This will reduce the effort for work power, time, place and cost, opening for a possibility of utilizing for many other areas such as linguistics study and public sector.

• Knowledge Management Research
• /
• v.10 no.4
• /
• pp.31-42
• /
• 2009

With the rapid development of information technology, e-learning is growing in corporate. However, there are still problems in learning, such as low learning persistence rate. Learning outcomes are complex phenomenon influenced by a multitude of factors, it is need to considering the direct and indirect causal relationship among various factors. Therefore, the purpose of this study was to develop the causal model that explains the learning outcomes (satisfaction learning persistence) in corporate e-learning. This study was also intended to examine the causal relationship between the interaction and learning persistence through satisfaction mediators. For this, online survey was conducted with a sample of 270 learners who enrolled in cyber training course at A company. The major findings of this study are as follows: First, internality (internal locus of control, ${\beta}=.154$), interaction (${\beta}=.489$), satisfaction (${\beta}=.304$) have direct effect on learning persistence. Second, the interaction has direct effect on the satisfaction (${\beta}=.320$). Third, the satisfaction has direct effect on the learning persistence, and mediating the interaction and learning persistence. This result will contribute to build a learning strategy to improve learning outcomes.

• Journal of Families and Better Life
• /
• v.21 no.3
• /
• pp.123-131
• /
• 2003

The potential of information and communication technologies has already had a profound impact on many ways of our living and society. To catch up with this trend, Korean government has demonstrated the vision for informatization called‘Cyber Korea 21'/'e-Korea’project and has fully supported the education for informatization which is one of the key factors to reduce the digital divide. But, In spite of developing a variety of training programs with different target groups and policy objectives, a digital divide remains in some cases even while Internet access and computer ownership are rising rapidly for almost all groups. For example, the noticeable divides still exists between men and women. To accelerate the process of women informatization, which we believe to be major contributors for the high quality of life for women, we need to promote full-time housewives to become aggressive information users and producers. Internet shopping, for example, might be a good starting point for motivating women to become active information users. In this paper, we present a model for explaining the factors affecting the degree of satisfaction of housewives from internet shopping. Based on data collected from a questionnaire survey from housewives in Seoul, we conclude that the perceived usefulness, ease of use and the playfulness significantly affect the level of satisfaction, but the playfulness does not directly affect the intention to revisit and purchase. In addition, we found out that the perceived usefulness is affected by efficiency, attitude and easy to access. We also provide other interesting statistical results and implications.

• Journal of the Korea Society of Computer and Information
• /
• v.18 no.8
• /
• pp.141-148
• /
• 2013

Traditional face-to-face university education has shifted its course to combine the advantages of both online and offline education in a blended-learning approach. However, there is still much that is unknown about the actual effect of blended learning, particularly it's learning outcomes in terms of cost effectiveness. This study qualitatively examines the costs and the learning outcomes of blended learning at an on-line college and off-line university. Online college level English pedagogy courses and blended with offline operations at an online university were studied across two semesters in terms of the quality of education, and both direct and indirect cost savings. Other causes for various learning outcomes and cost implications are proposed and validated.

• Journal of Korea Game Society
• /
• v.19 no.2
• /
• pp.111-122
• /
• 2019

This study investigated the psychological effects of using a serious game with young learners in the English classroom compared with those of a dictionary application. A tablet PC-based serious game was created for the training of English phoneme discrimination for Korean 6th graders, and its psychological effects were measured using a paper-based survey and face-to-face interviews. The overall results revealed that the serious game had more positive psychological effects on young learners than the dictionary app. These findings provide supporting empirical evidence for using serious games in English classrooms.