• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.20 no.10
• /
• pp.277-283
• /
• 2019

Advances in technology have allowed the development and commercialization of various parts; however this has shortened the discontinuation cycle of the components. This means that repair and logistic support of weapon system which is applied to thousands of part components and operated over the long-term is difficult, which is the one of main causes of the decrease in the availability of weapon system. To improve this problem, the United States has created a special organization for this problem, whereas in Korea, commercial tools are used to predict and manage DMSMS. However, there is rarely a method to predict life cycle of parts that are not presented DMSMS information at the commercial tools. In this study, the structured and unstructured data of parts of a commercial tool were gathered, preprocessed, and embedded using neural network algorithm. Then, a method is suggested to predict the life cycle risk (LC Risk) and year to end of life (YTEOL). In addition, to validate the prediction performance of LC Risk and YTEOL, the prediction value is compared with descriptive statistics.

• KIPS Transactions on Software and Data Engineering
• /
• v.8 no.11
• /
• pp.441-448
• /
• 2019

Previous researches on automatic spacing of Korean sentences has been researched to correct spacing errors by using n-gram based statistical techniques or morpheme analyzer to insert blanks in the word boundary. In this paper, we propose an end-to-end automatic word spacing by using deep neural network. Automatic word spacing problem could be defined as a tag classification problem in unit of syllable other than word. For contextual representation between syllables, Bi-LSTM encodes the dependency relationship between syllables into a fixed-length vector of continuous vector space using forward and backward LSTM cell. In order to conduct automatic word spacing of Korean sentences, after a fixed-length contextual vector by Bi-LSTM is classified into auto-spacing tag(B or I), the blank is inserted in the front of B tag. For tag classification method, we compose three types of classification neural networks. One is feedforward neural network, another is neural network language model and the other is linear-chain CRF. To compare our models, we measure the performance of automatic word spacing depending on the three of classification networks. linear-chain CRF of them used as classification neural network shows better performance than other models. We used KCC150 corpus as a training and testing data.

• Journal of the Korea institute for structural maintenance and inspection
• /
• v.25 no.6
• /
• pp.60-67
• /
• 2021

In lately, it's been developed and used a system of using deck plates as formwork in order to solve various problems caused by conventional formwork system. This system is more economical and has higher constructability than the conventional system by permanently embedding most of deck plates into the members. However, for this kind of embedded deck plates formwork system, it's been reported that it is difficult to verify filling of concrete in members like beams with narrow width and complicated rebar arrangement. In addtion, there are several problems such as corrosion of deck plates in terms of constructability and maintenance. Therefore, in this study, it is attempted to develop a removal-deck plate formwork system for beams by removing deck plates after concrete curing. The system consists of a deck plate module that acts as form, a frame preventing deformation by concrete lateral pressure, stirrup frame, and connector that combines these. As a result of this research, it is verified that it has higher constructability, efficiently prevents deformation caused by concrete lateral pressure and could be easily removed in the developed formwork system.

• The Journal of Society for e-Business Studies
• /
• v.23 no.3
• /
• pp.33-49
• /
• 2018

The importance of convergence activities among business is increasing due to the necessity of designing and developing new products to satisfy various customers' needs. In particular, decision makers such as CEOs are required to participate in networks between entrepreneurs for being connected with valuable convergence partners. Moreover, it is important for entrepreneurs not only to make a large number of network connections, but also to understand the networking relationship with entrepreneurs with similar topic information. However, there is a difficult limit in collecting the topic information that can show the lack of current status of business and the technology and characteristics of entrepreneur in industry sector. In this paper, we solve these problems through the topic extraction method and analyze the business network in three aspects. Specifically, there are C, S, T-Layer models, and each model analyzes amount of entrepreneurs relationship, network centrality, and topic similarity. As a result of experiments using real data, entrepreneur need to activate network by connecting high centrality entrepreneur when the corporate relationship is low. In addition, we confirmed through experiments that there is a need to activate the topic-based network when topic similarity is low between entrepreneurs.

• Smart Media Journal
• /
• v.8 no.2
• /
• pp.58-65
• /
• 2019

Automatic text summarization is a process of shortening a text document by either extraction or abstraction. The abstraction approach inspired by deep learning methods scaling to a large amount of document is applied in recent work. Abstractive text summarization involves utilizing pre-generated word embedding information. Low-frequent but salient words such as terminologies are seldom included to dictionaries, that are so called, out-of-vocabulary(OOV) problems. OOV deteriorates the performance of Encoder-Decoder model in neural network. In order to address OOV words in abstractive text summarization, we propose a copy mechanism to facilitate copying new words in the target document and generating summary sentences. Different from the previous studies, the proposed approach combines accurate pointing information and selective copy mechanism based on bidirectional RNN and bidirectional LSTM. In addition, neural network gate model to estimate the generation probability and the loss function to optimize the entire abstraction model has been applied. The dataset has been constructed from the collection of abstractions and titles of journal articles. Experimental results demonstrate that both ROUGE-1 (based on word recall) and ROUGE-L (employed longest common subsequence) of the proposed Encoding-Decoding model have been improved to 47.01 and 29.55, respectively.

• Journal of the Korea institute for structural maintenance and inspection
• /
• v.23 no.2
• /
• pp.84-91
• /
• 2019

Recently, there has been a tendency to improve seismic performance of building structure by installing a steel hysteretic damping device which is economically efficient and easy to install and maintain. However, for a reinforced concrete building, a set of complicated connecting hardware and braces to fix the steel hysteretic damping device yields deteriorated reliability in damping performance. Therefore, this study presents a method of directly embedding a Kagome damper, which was investigated in previous researches, into a concrete structure without additional connecting hardware. Moreover, in this study, a series of experiments conducted to provide a basis of the Kagome damper by confirming the seismic behavior for various embedded lengths. As a result, in a group of the embedded length of $1.0l_d$, the dampers were pulled out, while concrete breakout occurs. In a group of $2.0l_d$, neither pull-out nor concrete breakout occurred, while the dampers show stable behavior. Moreover, the buried length of $2.0l_d$ has 1.3 times better energy dissipation capacity. The system presented in this study can reduce the cost and period for installing, omitting making additional hardware.

• Science of Emotion and Sensibility
• /
• v.22 no.1
• /
• pp.89-100
• /
• 2019

Sleep apnea, a medical condition associated with a variety of complications, is generally monitored by standard sleep polysomnography, which is expensive and uncomfortable. To overcome these limitations, this study proposes an unconstrained wearable monitoring system with stretch-fiber sensors that integrate with the wearer's clothing. The system allows patients to undergo examinations in a familiar environment while minimizing the occurrence of skin allergies caused by adhesive tools. As smart clothing for adult males with sleep apnea, long-sleeved T-shirts embedding fibrous sensors were developed, enabling real-time monitoring of the patients' breathing rate, oxygen saturation, and airflow as sleep apnea diagnostic indicators. The gauge factor was measured as 20.3 in sample 4. The maximum breathing intake, measured during three large breaths, was 2048 ml. the oxygen saturation was measured before and during breath-holding. The oxygen saturation change was 69.45%, showing a minimum measurable oxygen saturation of 70%. After washing the garment, the gauge factor reduced only to 18.0, confirming the durability of the proposed system. The wearable sleep apnea monitoring smart clothes are readily available in the home and can measure three indicators of sleep apnea: respiration rate, breathing flow and oxygen saturation.

• The Journal of Internal Korean Medicine
• /
• v.40 no.1
• /
• pp.13-37
• /
• 2019

Objectives: The purpose of this study was to assess the effects of Korean medicine for non-alcoholic fatty liver disease (NAFLD). We analyzed the result of randomized controlled trials (RCTs) that applied Korean medicine to NAFLD patients through meta analysis and systematic review. Methods: The key question was to the effects of Korean medicine for NAFLD patients according to the PICO-SD (participants, intervention, comparison, outcome, study design) and we included only RCTs. We searched 10 databases including NDSL, KMBASE, KISS, KISTI, KoreaMed, Koreantk, OASIS, Cochrane, Pubmed, EMBASE without a language restriction. We assessed risk of bias by Cochrane group's Risk of Bias tool. Results: The finally selected 12 RCTs were analyzed. Total number of participants was 1189 (male 719, female 470) as 684 and 505 in the oriental medicine group (i.e. herbal drugs, acupuncture, acupoint embedding therapy) and control group (conventional drugs, placebo), respectively. The meta analysis results of examining 7 RCTs comparing the therapeutic efficacy of herbal medicine with that of Western medicine showed statistically significant (p<0.05) differences in the efficacy evaluation, liver function test results, blood lipids, and TNF-${\alpha}$. Furthermore, the meta analysis results of investigating 3 RCTs comparing the therapeutic efficacy of herbal medicine with that of a placebo showed statistically significant (p<0.05) differences in the liver function test results, blood lipids, and waist circumference. Conclusions: The research showed that Korean medicine for NAFLD can be effective treatment. But more studies are required to enhance the level of evidence and we should report on safety.

• Journal of the Korea Computer Graphics Society
• /
• v.27 no.3
• /
• pp.65-74
• /
• 2021

3D reconstruction means that reconstructing the 3D shape of the object in an image and a video. We proposed a progressive occupancy network architecture that can recover not only the overall shape of the object but also the local details. Unlike the original occupancy network, which uses a feature vector embedding information of the whole image, we extract and utilize the different levels of image features depending on the receptive field size. We also propose a novel network architecture that applies the image features sequentially to the decoder blocks in the decoder and improves the quality of the reconstructed 3D shape progressively. In addition, we design a novel decoder block structure that combines the different levels of image features properly and uses them for updating the input point feature. We trained our progressive occupancy network with ShapeNet. We compare its representation power with two prior methods, including prior occupancy network(ONet) and the recent work(DISN) that used different levels of image features like ours. From the perspective of evaluation metrics, our network shows better performance than ONet for all the metrics, and it achieved a little better or a compatible score with DISN. For visualization results, we found that our method successfully reconstructs the local details that ONet misses. Also, compare with DISN that fails to reconstruct the thin parts or occluded parts of the object, our progressive occupancy network successfully catches the parts. These results validate the usefulness of the proposed network architecture.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.14 no.5
• /
• pp.907-916
• /
• 2019

In this paper, we propose a new five-qubit multiple bit flip code that can completely protect the target qubit from all multiple bit flip errors using only CNOT gates. The proposed multiple bit flip codes can be easily extended to multiple phase flip codes by embedding Hadamard gate pairs in the root error section as in conventional single bit flip code. The multiple bit flip code and multiple phase flip code in this paper share the state vector error information by four auxiliary qubits. These four-qubit state vectors reflect the characteristic that all the multiple flip errors with Pauli X and Z corrections commonly include a specific root error. Using this feature, this paper shows that low-cost implementation is possible despite the QECC design for multiple-flip error correction by batch processing the detection and correction of Pauli X and Z root errors with only three CNOT gates. The five-qubit multiple bit flip code and multiple phase flip code proposed in this paper have 100% error correction rate and 50% error discrimination rate. All QECCs presented in this paper were verified using QCAD simulator.