• Knowledge Management Research
• /
• v.15 no.2
• /
• pp.23-41
• /
• 2014

Firms increasingly utilize Social Networking Service(SNS) to lead user's voluntary behavior. In the web-based environment, users show coexist loyal behavior which is represented by 'web-based pro-organization citizenship behavior' and 'anti-citizenship behavior'. To measure genuine performance of web-activity, we separated degree of compliance based on credibility, 'deep-level' and 'surface-level' to comprehend different behavior after compliance. The analysis result shows that contents credibility is important to enhance deep-level of compliance which has significant influence on web-based pro-organization citizenship behavior. Contrastively, surface-level of compliance has influence on anti-citizenship behavior. Based on the results of these analyses, the directions of web-based activities for the common good and self-interests of the stakeholders of the web-based activities will be proposed.

• Journal of the Korea Society of Computer and Information
• /
• v.23 no.11
• /
• pp.31-41
• /
• 2018

Text classification has been studied for a long time in the Natural Language Processing field. In this paper, we propose an article- and paragraph-level genre classification system using Word2Vec-based LSTM, GRU, and CNN models for large-scale English corpora. Both article- and paragraph-level classification performed best in accuracy with LSTM, which was followed by GRU and CNN in accuracy performance. Thus, it is to be confirmed that in evaluating the classification performance of LSTM, GRU, and CNN, the word sequential information for articles is better than the word feature extraction for paragraphs when the pre-trained Word2Vec-based word embeddings are used in both deep learning-based article- and paragraph-level classification tasks.

• Safety and Health at Work
• /
• v.9 no.4
• /
• pp.441-446
• /
• 2018

Background: The emotional labor performed by organization members affects psychological well-being at the individual level, which consequently affects results at the organizational level. Moreover, despite evidence that the customer orientation and service level of nurses greatly affect hospital management, studies that comprehensively analyze emotional labor, work burnout, and work engagement related to customer orientation and service level are lacking. This study investigated relationships and paths by designing a model of the effect of emotional labor performed by nurses on the level of service delivery and customer orientation. Methods: This survey-based study was based on a path analysis designed to verify a hypothesized model involving emotional labor performed by nurses, level of service delivery, customer orientation, work engagement, and burnout. Questionnaires were distributed to 378 nurses in general hospitals with more than 500 beds located in Seoul, Republic of Korea, between March 25 and April 8, 2013. Results: The results showed that deep acting and work engagement had direct and indirect effects on increasing the level of service delivery and customer orientation of nurses. However, surface acting had an indirect effect on reducing the level of service delivery and customer orientation. Conclusion: It would be more effective to develop interventions to enhance deep acting and work engagement than to attempt to reduce surface acting and work burnout in clinical nursing settings.

• Proceedings of the KIEE Conference
• /
• 2011.07a
• /
• pp.1702-1703
• /
• 2011

MEMS 소자의 Wafer level Package 관련하여 Deep cavity를 가진 Cap Wafer와 Polymer bonding 중 cavity 단차로 인한 Polymer Patterning 및 접합 불량의 어려움을 극복할 수 있는 새로운 공정 flow를 제안하였다. Cavity를 형성할 때 사용하는 Si deep etching Mask인 기존의 Photoresist를 접합용 감광성 Polymer로 대체하고, cavity 형성 후, 별도의 추가 공정 없이 이 Polymer를 이용해 Wafer bonding을 진행하였다. 이를 통해 cavity 단차에 따른 문제를 해결함과 동시에 공정이 단순하고 제작 비용이 저렴하며, 신뢰성 있는 Wafer level Package를 구현하였다.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.34 no.5C
• /
• pp.485-497
• /
• 2009

In this paper, we described DVB-S2 system for mobility. cross layer coding technique are needed to maintain the performance in deep fading channel. Cross layer coding is divided into two kinds of level. First level is Physical layer coding and, second layer is link layer or upper layer coding. Fixed on DVB-S2 short frame coding method as a physical layer, we simulated the various coding method as an upper layer coding. Furthermore, we analyzed the performance of each coding method on according to mobile vehicle speed, data rate, interleaving memory size, and IP packet size.

• The Journal of Korean Physical Therapy
• /
• v.31 no.1
• /
• pp.1-6
• /
• 2019

Purpose: This paper proposes proper and effective neck exercises by comparing the deep and superficial cervical flexor muscle activities and thickness according to the pressure level during cranio-cervical flexion exercises between a normal posture group and forward head posture group. Methods: A total of 20 subjects (8 males and 12 females) without neck pain and disabilities were selected. The subjects' craniovertebral angles were measured; they were divided into a normal posture and a forward head posture group. During cranio-cervical flexion exercises, the thickness of the deep cervical flexor neck muscle and the activity of the surface neck muscles were measured using ultrasound and EMG. Results: The results showed that the thickening of the deep cervical flexor was increased significantly to 28 and 30 mmHg in the forward head posture group. The sternocleidomastoid muscle activity increased significantly to 24, 26, 28, and 30 mmHg in the forward head posture group. The anterior scalene muscle activity increased significantly to 26, 28, and 30mmHg in the forward head posture group. A significant difference of 26, 28, and 30 mmHg in the sternocleidomastoid and anterior scalene muscles was observed between two groups. Conclusion: To prevent a forward head posture and maintain proper cervical curve alignment, the use of the superficial cervical flexor muscles must be minimized. In addition, to perform a cranio-cervical flexion exercises to effectively activate the deep cervical flexor muscles, 28 and 30 mmHg for normal posture adults and 28 mmHg for adults with forward head postures are recommended.

• ETRI Journal
• /
• v.44 no.3
• /
• pp.462-475
• /
• 2022

Although researchers have proposed numerous techniques for speech emotion recognition, its performance remains unsatisfactory in many application scenarios. In this study, we propose a speech emotion recognition model based on a genetic algorithm (GA)-decision tree (DT) fusion of deep and acoustic features. To more comprehensively express speech emotional information, first, frame-level deep and acoustic features are extracted from a speech signal. Next, five kinds of statistic variables of these features are calculated to obtain utterance-level features. The Fisher feature selection criterion is employed to select high-performance features, removing redundant information. In the feature fusion stage, the GA is is used to adaptively search for the best feature fusion weight. Finally, using the fused feature, the proposed speech emotion recognition model based on a DT support vector machine model is realized. Experimental results on the Berlin speech emotion database and the Chinese emotion speech database indicate that the proposed model outperforms an average weight fusion method.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.15 no.12
• /
• pp.4456-4475
• /
• 2021

The system for re-identifying persons is used to find and verify the persons crossing through different spots using various cameras. Much research has been done to re-identify the person by utilising features with deep-learned or hand-crafted information. Deep learning techniques segregate and analyse the features of their layers in various forms, and the output is complex feature vectors. This paper proposes a distinctive framework called Integrated Level Feature Pattern (ILFP) framework, which integrates local and global features. A new deep learning architecture named modified XceptionNet (m-XceptionNet) is also proposed in this work, which extracts the global features effectively with lesser complexity. The proposed framework gives better performance in Rank1 metric for Market1501 (96.15%), CUHK03 (82.29%) and the newly created NEC01 (96.66%) datasets than the existing works. The mean Average Precision (mAP) calculated using the proposed framework gives 92%, 85% and 98%, respectively, for the same datasets.

• Asia pacific journal of information systems
• /
• v.33 no.3
• /
• pp.737-767
• /
• 2023

With the availability of big customer data and advances in machine learning techniques, the prediction of customer behavior at the session-level has attracted considerable attention from marketing practitioners and scholars. This study aims to predict customer purchase conversion at the session-level by employing customer profile, transaction, and clickstream data. For this purpose, we develop a multimodal deep learning fusion model with dynamic and static features (i.e., DS-fusion). Specifically, we base page views within focal visist and recency, frequency, monetary value, and clumpiness (RFMC) for dynamic and static features, respectively, to comprehensively capture customer characteristics for buying behaviors. Our model with deep learning architectures combines these features for conversion prediction. We validate the proposed model using real-world e-commerce data. The experimental results reveal that our model outperforms unimodal classifiers with each feature and the classical machine learning models with dynamic and static features, including random forest and logistic regression. In this regard, this study sheds light on the promise of the machine learning approach with the complementary method for different modalities in predicting customer behaviors.

• Journal of Information Processing Systems
• /
• v.15 no.1
• /
• pp.91-99
• /
• 2019

An innovative tone modeling framework based on deep neural networks in tone recognition was proposed in this paper. In the framework, both the prosodic features and the articulatory features were firstly extracted as the raw input data. Then, a 5-layer-deep deep belief network was presented to obtain high-level tone features. Finally, support vector machine was trained to recognize tones. The 863-data corpus had been applied in experiments, and the results show that the proposed method helped improve the recognition accuracy significantly for all tone patterns. Meanwhile, the average tone recognition rate reached 83.03%, which is 8.61% higher than that of the original method.