• IEMEK Journal of Embedded Systems and Applications
• /
• v.15 no.2
• /
• pp.43-49
• /
• 2020

In this paper, we propose a stage-wise knowledge transfer method that uses block-wise retraining to transfer the useful knowledge of a pre-trained residual network (ResNet) in a teacher-student framework (TSF). First, multiple hint information transfer and block-wise supervised retraining of the information was alternatively performed between teacher and student ResNet models. Next, Softened output information-based knowledge transfer was additionally considered in the TSF. The results experimentally showed that the proposed method using multiple hint-based bottom-up knowledge transfer coupled with incremental block-wise retraining provided the improved student ResNet with higher accuracy than existing KD and hint-based knowledge transfer methods considered in this study.

• The Journal of Korean Association of Computer Education
• /
• v.18 no.6
• /
• pp.71-79
• /
• 2015

In a new era of Cloud and Big data, how to search the useful data from dynamic huge data pool in a right time and right way is most important at the stage where the information is getting more important. Above all, in the era of s Big Data it is required to design the advanced efficient intelligent Knowledge system which can process the dynamic variable big data. Accordingly in this paper we propose Dynamic personal Knowledge Network as one of the advanced Intelligent system approach. Adopting the human brain function and its neuro dynamics, an Intelligent system which has a structural flexibility was designed. For Structure-Function association, a personal Knowledge Network is made to be structured and to have reorganizing function as connecting the common nodes. We also design this system to have a reasoning process in the extracted optimal paths from the Knowledge Network.

• Journal of the Korean Operations Research and Management Science Society
• /
• v.19 no.2
• /
• pp.129-151
• /
• 1994

Recently, Decision Support Systems (DSS) research has seen a more to combine Artificial Intelligence (AI) including neural network techniques with traditional DSS concepts and technologies to build an intelligent DSS or a knowledge-based DSS. This article proposes a Management Evaluation and its Support System (MESS) as a knowledge-based DSS. The management evaluation of a firm means the performance of all managerial operations is appraised by considering the situations of the firm. A neural network is used to represent the management evaluation structure as a suitable means of management knowledge representation. Finally a case study in a telecommunication corporation is presented.

• Journal of the Korea Society of Computer and Information
• /
• v.23 no.10
• /
• pp.173-180
• /
• 2018

In this paper, we analyzed key keywords and research themes in the field of defense research using keyword network analysis and tried to grasp the whole knowledge structure. To do this, we extracted data from 2,165 research data from defense related research institutes from 2010 to 2017 and applied the Pareto rule to the number of abstracts of words and the number of links between words, We extracted a total of 2,303 words based on the criterion and extracted 204 final key words through component analysis. By analyzing the centrality and cohesiveness through these key words, we confirmed the concept of core research in the defense field and derived a total of 7 large groups and 16 small groups of each group in the knowledge structure of the defense area.

• Knowledge Management Research
• /
• v.22 no.4
• /
• pp.41-70
• /
• 2021

The use of social media has many advantages such as knowledge sharing, social networking, and communicating with other people. However, it has given rise to various side effects including stress, Which is defined as social network stress in this study. This study aims to conceptualize social network stress and investigate its effect on switching behavior in social media. For this purpose, we present a research model that consists of the antecedents and consequences of social network stress and test it empirically using LISREL 8.7 based on the structural equation model. The empirical results showed that knowledge sharing and self-disclosure had positive impact on social network stress, which in turn positively influenced social media switching behaviors. In conclusion, we discussed both theoretical and practical implications of this research and suggested its limitations.

• Knowledge Management Research
• /
• v.17 no.4
• /
• pp.77-104
• /
• 2016

The effective management of R&D employees is critical for a small or medium sized firm's sustainable growth. R&D employees have professional skills and choose expertise-oriented or management-oriented careers in the process of organizational socialization. This study synthetically verifies the direct and indirect effects of R&D employees' social network and self-esteem on their career orientation by organizational commitment based on social network theory and social recognition theory. The research model has been analyzed through structural equation modeling using survey responses from 220 R&D employees at small- and medium-sized firms in Korea. The analysis results show that internal network activities have direct and indirect impacts on organizational commitment and career orientation, but external network activities do not have significant effects on self-esteem, organizational commitment, or career orientation. There is no consensus in prior studies on whether expert orientation and management orientation are distinct concepts. In this study, these two types of orientation are verified as distinct concepts. It is also found that R&D employees' internal network activities are significant factors for a company's growth. A company should implement an educational system of roles and duties using which individuals can pursue career progression. In addition, it is necessary to provide career development programs such as job rotation, mentoring, and career counseling.

• Asian Journal of Innovation and Policy
• /
• v.4 no.2
• /
• pp.242-262
• /
• 2015

The relationship between geographical proximity and academics' formal and informal knowledge-transfer activities in the network is analyzed with a mixed research method. With social network analysis as a basis, we have explored the networks between academics and firms in the 16 regions of South Korea. The result shows Seoul and Gyunggi are identified as central nodes, meaning that the academics in other regions tend to collaborate with firms in these regions. An econometric analysis is performed to confirm the localization of knowledge-transfer activities. The intensity of formal channels measured by the number of academic papers is negatively, but significantly associated with the geographical proximity. However, we have not found any significant relationship between the formality of the channels and geographical proximity. Possibly, the regional innovation systems in South Korea are neither big enough nor strong enough to show a localization effect.

• Proceedings of the Korean Society for Bioinformatics Conference
• /
• 2005.09a
• /
• pp.366-370
• /
• 2005

Tissue microarry is one of the high throughput technologies in the post-genomic era. Using tissue microarray, the researchers are able to investigate large amount of gene expressions at the level of DNA, RNA, and protein The important aspect of tissue microarry is its ability to assess a lot of biomarkers which have been used in clinical practice. To manipulate the categorical data of tissue microarray, we applied Bayesian network classifier algorithm. We identified that Bayesian network classifier algorithm could analyze tissue microarray data and integrating prior knowledge about gastric cancer could achieve better performance result. The results showed that relevant integration of prior knowledge promote the prediction accuracy of survival status of the immunohistochemical tissue microarray data of 18 tumor suppressor genes. In conclusion, the application of Bayesian network classifier seemed appropriate for the analysis of the tissue microarray data with clinical information.

• Journal of Korea Multimedia Society
• /
• v.13 no.6
• /
• pp.890-900
• /
• 2010

The Bayesian networks methods provide an efficient tool for performing information fusion and decision making under conditions of uncertainty. This paper proposes Bayes estimators for the system effectiveness in energy saving of the wireless sensor networks by use of the Bayesian method under the non-informative prior knowledge about means of active and sleep times based on time frames of sensor nodes in a wireless sensor network. And then, we conduct a case study on some Bayesian estimation models for the system energy saving effectiveness of a wireless sensor network, and evaluate and compare the performance of proposed Bayesian estimates of the system effectiveness in energy saving of the wireless sensor network. In the case study, we have recognized that the proposed Bayesian system energy saving effectiveness estimators are excellent to adapt in evaluation of energy efficiency using non-informative prior knowledge from previous experience with robustness according to given values of parameters.

• Asia pacific journal of information systems
• /
• v.16 no.4
• /
• pp.203-215
• /
• 2006

The advent of KM(knowledge management) concept have led many organizations to seek an effective way to make use of their knowledge. But the absence of right tools for systematic handling of unstructured information makes it difficult to automatically retrieve and share relevant information that exactly meet user's needs. we propose a systematic method to enable content-based information retrieval from corpus of unstructured documents. In our method, a document is represented by using several key terms which are automatically selected based on their quantitative relevancy to the document. Basically, the relevancy is calculated by using a traditional TFIDF measure that are widely accepted in the related research, but to improve effectiveness of the measure, we exploited 'concept network' that represents term-term relationships. In particular, in constructing the concept network, we have also considered relative position of terms occurring in a document. A prototype system for experiment has been implemented. The experiment result shows that our approach can have higher performance over the conventional TFIDF method.