• Journal of The Korean Association of Information Education
• /
• v.26 no.3
• /
• pp.197-207
• /
• 2022

In this paper, we develop a machine learning based block code generation and recommendation model for the purpose of reducing cognitive load of learners during coding education that learns the learner's block that has been made in the block programming environment using natural processing model and fine-tuning and then generates and recommends the selectable blocks for the next step. To develop the model, the training dataset was produced by pre-processing 50 block codes that were on the popular block programming language web site 'Entry'. Also, after dividing the pre-processed blocks into training dataset, verification dataset and test dataset, we developed a model that generates block codes based on LSTM, Seq2Seq, and GPT-2 model. In the results of the performance evaluation of the developed model, GPT-2 showed a higher performance than the LSTM and Seq2Seq model in the BLEU and ROUGE scores which measure sentence similarity. The data results generated through the GPT-2 model, show that the performance was relatively similar in the BLEU and ROUGE scores except for the case where the number of blocks was 1 or 17.

• Journal of the Korea Convergence Society
• /
• v.12 no.11
• /
• pp.207-220
• /
• 2021

Recent advances in new media for sensory experiences keep expanding visual expression methods in installation art such as projection mapping and virtual reality. Artists can create and develop visual expression techniques based on such new media. Projection mapping is a new medium that continues to add various possibilities to visual expression in media art. Under the projection mapping environment, artists can recompose the object or space with the digital content by projecting video onto three-dimensional surfaces in the space. This paper focuses on the process where visual expression with the projection mapping technology leads to viewers' sensory experience. To this end, "reproducibility," "dissemination," "virtuality," and "interactivity" of media were analyzed to describe the meaning and *definition of visual expression. Artworks are considered as an example to study visual expression techniques such as "repetition and overlap," "simulacrum and metaphor," and "displacement and conversion." I applied the analysis and created Domino, a projection mapping artwork, which helps the research on visual expression techniques that can lead to sensory experience the extensibility of visual expression.

• Journal of the Korean Electrochemical Society
• /
• v.26 no.1
• /
• pp.11-18
• /
• 2023

As the use of lithium-ion secondary batteries is rapidly increasing due to the rapid growth of the electric vehicle market, the disposal and recycling of spent batteries after use has been raised as a serious problem. Since stored energy must be removed in order to recycle the spent batteries, an effective discharging process is required. In this study, graphite and NCM622 were used as active materials to manufacture coin-type half cells and full cells, and the electrochemical behavior occurring during overdischarge was analyzed. When the positive and negative electrodes are overdischarged respectively using a half-cell, a conversion reaction in which transition metal oxide is reduced to metal occurs first in the positive electrode, and a side reaction in which Cu, the current collector, is corroded following decomposition of the SEI film occurs in the negative electrode. In addition, a side reaction during overdischarge is difficult to occur because a large polarization at the initial stage is required. When the full cell is overdischarged, the cell reaches 0 V and the overdischarge ends with almost no side reaction due to this large polarization. However, if the full cell whose capacity is degraded due to the cycle is overdischarged, corrosion of the Cu current collector occurs in the negative electrode. Therefore, cycled cell requires an appropriate treatment process because its electrochemical behavior during overdischarge is different from that of a fresh cell.

• Journal of Korean Society of Forest Science
• /
• v.112 no.3
• /
• pp.267-279
• /
• 2023

This research assessed the feasibility of using high-resolution aerial images and deep learning algorithms for estimating the land-use and land-cover areas at the Approach 3 level, as outlined by the Intergovernmental Panel on Climate Change. The results from different sampling densities of high-resolution (51 cm) aerial images were compared with the land-cover map, provided by the Ministry of Environment, and analyzed to estimate the accuracy of the land-use and land-cover areas. Transfer learning was applied to the VGG16 architecture for the deep learning model, and sampling densities of 4 × 4 km, 2 × 4 km, 2 × 2 km, 1 × 2 km, 1 × 1 km, 500 × 500 m, and 250 × 250 m were used for estimating and evaluating the areas. The overall accuracy and kappa coefficient of the deep learning model were 91.1% and 88.8%, respectively. The F-scores, except for the pasture category, were >90% for all categories, indicating superior accuracy of the model. Chi-square tests of the sampling densities showed no significant difference in the area ratios of the land-cover map provided by the Ministry of Environment among all sampling densities except for 4 × 4 km at a significance level of p = 0.1. As the sampling density increased, the standard error and relative efficiency decreased. The relative standard error decreased to ≤15% for all land-cover categories at 1 × 1 km sampling density. These results indicated that a sampling density more detailed than 1 x 1 km is appropriate for estimating land-cover area at the local level.

• Journal of KIISE:Software and Applications
• /
• v.31 no.10
• /
• pp.1332-1344
• /
• 2004

Upon development of information super-highway and multimedia-related technoiogies in recent years, more efficient technologies to transmit, store and retrieve the multimedia data are required. Among such technologies, firstly, it is common that the semantic-based image retrieval is annotated separately in order to give certain meanings to the image data and the low-level property information that include information about color, texture, and shape Despite the fact that the semantic-based information retrieval has been made by utilizing such vocabulary dictionary as the key words that given, however it brings about a problem that has not yet freed from the limit of the existing keyword-based text information retrieval. The second problem is that it reveals a decreased retrieval performance in the content-based image retrieval system, and is difficult to separate the object from the image that has complex background, and also is difficult to extract an area due to excessive division of those regions. Further, it is difficult to separate the objects from the image that possesses multiple objects in complex scene. To solve the problems, in this paper, I established a content-based retrieval system that can be processed in 5 different steps. The most critical process of those 5 steps is that among RGB images, the one that has the largest and the smallest background are to be extracted. Particularly. I propose the method that extracts the subject as well as the background by using an Image, which has the largest background. Also, to solve the second problem, I propose the method in which multiple objects are separated using RGB channel selection techniques having optimized the excessive division of area by utilizing Watermerge's threshold value with the object separation using the method of RGB channels separation. The tests proved that the methods proposed by me were superior to the existing methods in terms of retrieval performances insomuch as to replace those methods that developed for the purpose of retrieving those complex objects that used to be difficult to retrieve up until now.

• Radiation Oncology Journal
• /
• v.25 no.1
• /
• pp.7-15
• /
• 2007

[ $\underline{Purpose}$ ]: To determine the patterns of evaluation and treatment in patients with breast cancer after mastectomy and treated with radiotherapy. A nationwide study was performed with the goal of improving radiotherapy treatment. $\underline{Materials\;and\;Methods}$: A web- based database system for the Korean Patterns of Care Study (PCS) for 6 common cancers was developed. Randomly selected records of 286 eligible patients treated between 1998 and 1999 from 17 hospitals were reviewed. $\underline{Results}$: The ages of the study patients ranged from 20 to 80 years (median age 44 years). The pathologic T stage by the AJCC was T1 in 9.7% of the cases, T2 in 59.2% of the cases, T3 in 25.6% of the cases, and T4 in 5.3% of the cases. For analysis of nodal involvement, N0 was 7.3%, N1 was 14%, N2 was 38.8%, and N3 was 38.5% of the cases. The AJCC stage was stage I in 0.7% of the cases, stage IIa in 3.8% of the cases, stage IIb in 9.8% of the cases, stage IIIa in 43% of the cases, stage IIIb in 2.8% of the cases, and IIIc in 38.5% of the cases. There were various sequences of chemotherapy and radiotherapy after mastectomy. Mastectomy and chemotherapy followed by radiotherapy was the most commonly performed sequence in 47% of the cases. Mastectomy, chemotherapy, and radiotherapy followed by additional chemotherapy was performed in 35% of the cases, and neoadjuvant chemoradiotherapy was performed in 12.5% of the cases. The radiotherapy volume was chest wall only in 5.6% of the cases. The volume was chest wall and supraclavicular fossa (SCL) in 20.3% of the cases; chest wall, SCL and internal mammary lymph node (IMN) in 27.6% of the cases; chest wall, SCL and posterior axillary lymph node in 25.9% of the cases; chest wall, SCL, IMN, and posterior axillary lymph node in 19.9% of the cases. Two patients received IMN only. The method of chest wall irradiation was tangential field in 57.3% of the cases and electron beam in 42% of the cases. A bolus for the chest wall was used in 54.8% of the tangential field cases and 52.5% of the electron beam cases. The radiation dose to the chest wall was $45{\sim}59.4\;Gy$ (median 50.4 Gy), to the SCL was $45{\sim}59.4\;Gy$ (median 50.4 Gy), and to the PAB was $4.8{\sim}38.8\;Gy$, (median 9 Gy) $\underline{Conclusion}$: Different and various treatment methods were used for radiotherapy of the breast cancer patients after mastectomy in each hospital. Most of treatment methods varied in the irradiation of the chest wall. A separate analysis for the details of radiotherapy planning also needs to be followed and the outcome of treatment is needed in order to evaluate the different processes.

• Journal of Intelligence and Information Systems
• /
• v.28 no.2
• /
• pp.127-146
• /
• 2022

Recently, as word embedding has shown excellent performance in various tasks of deep learning-based natural language processing, researches on the advancement and application of word, sentence, and document embedding are being actively conducted. Among them, cross-language transfer, which enables semantic exchange between different languages, is growing simultaneously with the development of embedding models. Academia's interests in vector alignment are growing with the expectation that it can be applied to various embedding-based analysis. In particular, vector alignment is expected to be applied to mapping between specialized domains and generalized domains. In other words, it is expected that it will be possible to map the vocabulary of specialized fields such as R&D, medicine, and law into the space of the pre-trained language model learned with huge volume of general-purpose documents, or provide a clue for mapping vocabulary between mutually different specialized fields. However, since linear-based vector alignment which has been mainly studied in academia basically assumes statistical linearity, it tends to simplify the vector space. This essentially assumes that different types of vector spaces are geometrically similar, which yields a limitation that it causes inevitable distortion in the alignment process. To overcome this limitation, we propose a deep learning-based vector alignment methodology that effectively learns the nonlinearity of data. The proposed methodology consists of sequential learning of a skip-connected autoencoder and a regression model to align the specialized word embedding expressed in each space to the general embedding space. Finally, through the inference of the two trained models, the specialized vocabulary can be aligned in the general space. To verify the performance of the proposed methodology, an experiment was performed on a total of 77,578 documents in the field of 'health care' among national R&D tasks performed from 2011 to 2020. As a result, it was confirmed that the proposed methodology showed superior performance in terms of cosine similarity compared to the existing linear vector alignment.

• Journal of Nuclear Fuel Cycle and Waste Technology(JNFCWT)
• /
• v.12 no.3
• /
• pp.217-233
• /
• 2014

Securing the radiological safety is a prerequisite for the safe management of the naturally occurring radioactive materials (NORM) which cannot be reused. This becomes a crucial focus of our R&D efforts upon the implementation of the Act on Protective Action Guidelines against Radiation in the Natural Environment. To secure the safety, the establishment of technical bases and procedures for securing radiological safety related to the disposal of NORM is required. Thus, it is necessary to analyze the characteristics, to collect the data, to have the radiological safety assessment methodologies and tools, to investigate disposal methods and facilities, and to study the effects of the input data on the safety for the NORM wastes. Here, we assess the environmental impact of the NORM waste disposal with respect to the major domestic and foreign NORM characteristics. The data associated with major industries are collected/analyzed and the status of disposal facilities and methodologies relevant to the NORM wastes is investigated. We also suggest the conceptual design concept of a landfill disposal facility and the management plan with respect to the major NORM wastes characteristics. The radionuclide pathways are identified for the atmospheric transport and leachate release and the environmental impact assessment methodology for the NORM waste disposal is established using a relevant code. The assessment and analysis on the exposure doses and excessive cancer risks for the NORM waste disposal are performed using the characteristics of the representative domestic NORM wastes including flying ash, phosphor gypsum, and redmud. The results show that the exposure dose and the excessive cancer risks are very low to consider any radiation effects. This study will contribute to development in the areas of the regulatory technology for securing radiological safety relevant to NORM waste disposal and to the implementation technology for the Act.

• Journal of radiological science and technology
• /
• v.36 no.4
• /
• pp.327-335
• /
• 2013

This study investigates the case of clinical application for TomoDirect 3D-CRT(TD-3D) and TomoHelical 3D-CRT(TH-3D) with evaluating dose distribution for clinical application in each case. Treatment plans were created for 8 patients who had 3 dimensional conformal radiation therapy using TD-3D and TH-3D mode. Each patients were treated for sarcoma, CSI(craniospinal irradiaion), breast, brain, pancreas, spine metastasis, SVC syndrome and esophagus. DVH(dose volume histogram) and isodose curve were used for comparison of each treatment modality. TD-3D shows better dose distribution over the irradiation field without junction effect because TD-3D was not influenced by target length for sarcoma and CSI case. In breast case, dosimetric results of CTV, the average value of D 99%, D 95% were $49.2{\pm}0.4$ Gy, $49.9{\pm}0.4$ Gy and V 105%, V 110% were 0%, respectively. TH-3D with the dosimetric block decreased dose of normal organ in brain, pancreas, spine metastasis case. SCV syndrome also effectively decreased dose of normal organ by using dose block to the critical organs(spinal cord <38 Gy). TH-3D combined with other treatment modalities was possible to boost irradiation and was total dose was reduced to spinal cord in esophagus case(spinal cord <45 Gy, lung V 20 <20%). 3D-CRT using Tomotherapy could overcomes some dosimetric limitations, when we faced Conventional Linac based CRT and shows clinically proper dose distribution. In conclusion, 3D-CRT using Tomotherapy will be one of the effective 3D-CRT techniques.

• Horticultural Science & Technology
• /
• v.28 no.3
• /
• pp.449-455
• /
• 2010

'Moulinrouge' was selected as the best regenerating cultivar among 18 different spray-type chrysanthemum cultivars bred in the Gyeongnam Flowers Breeding Research Institute. When the leaf explants from standard- and spray-type chrysanthemum 'Jinba' and 'Moulinrouge' were incubated on MS basal medium supplemented with $0.5mg{\cdot}L^{-1}$ BA and $1.0mg{\cdot}L^{-1}$ NAA, both 'Jinba' and 'Moulinrouge' induced adventitious shoots that can be regenerated into plantlets. Based on these regeneration conditions, we developed an efficient $Agrobacterium$-mediated chrysanthemum 'Moulinrouge' transformation method by using sequential selection of shoots from low ($10mg{\cdot}L^{-1}$) to high ($30mg{\cdot}L^{-1}$) concentrations of kanamycin after co-cultivation of leaf explants with $Agrobacterium$ for 10 days and induction of shoots. All kanamycin resistant plants investigated with genomic PCR analysis carried the report gene, $AtSICKLE$, in their genome. Although expression levels of the report gene in the transgenic plants investigated with RT-PCR were relatively low because of inefficiency of CaMV 35S promoter in chrysanthemum, transgenic lines expressing $AtSICKLE$ efficiently showed leaf epinasty phenotype. We expect that our results will provide a useful method that can perform a high-throughput investigation of genes isolated and studied well in model plants for molecular breeding of chrysanthemum.