This study proposes a new collaborative filtering model that integrates Restricted Boltzmann Machines. The proposed two-stage model is applied to household-level supermarket purchase data. Results show that our model fits the data better and outperforms existing collaborative filtering methods in predicting shopping patterns. The proposed model also improves interpretations of market complexity and common causes of coincidence associated with customers’ multi-category purchases.

Natural uranium-contaminated soil in Korea Atomic Energy Research Institute (KAERI) was generated by decommissioning of the natural uranium conversion facility in 2010. Some of the contaminated soil was expected to be clearance level, however the disposal cost burden is increasing because it is not classified in advance. In this study, pre-classification method is presented according to the ratio of naturally occurring radioactive material (NORM) and contaminated uranium in the soil. To verify the validity of the method, the verification of the uranium radioactivity concentration estimation method through γ-ray analysis results corrected by self-absorption using MCNP6.2, and the validity of the pre-classification method according to the net peak area ratio were evaluated. Estimating concentration for 238U and 235U with γ-ray analysis using HPGe (GC3018) and MCNP6.2 was verified by 􀟙-spectrometry. The analysis results of different methods were within the deviation range. Clearance screening factors (CSFs) were derived through MCNP6.2, and net peak area ratio were calculated at 295.21 keV, 351.92 keV(214Pb), 609.31 keV, 1120.28 keV, 1764.49 keV(214Bi) of to the 92.59 keV. CSFs for contaminated soil and natural soil were compared with U/Pb ratio. CSFs and radioactivity concentrations were measured, and the deviation from the 60 minute measurement results was compared in natural soil. Pre-classification is possible using by CSFs measured for more than 5 minutes to the average concentration of 214Pb or 214Bi in contaminated soil. In this study, the pre-classification method of clearance determination in contaminated soil was evaluated, and it was relatively accurate in a shorter measurement time than the method using the concentrations. This method is expected to be used as a simple pre-classification method through additional research.

Since high-level radioactive wastes contain long-lived nuclides and emit high energy, they should be disposed of permanently through a deep geological disposal system. In Korea, the first (2016.07) and the second (2021.12) basic plans for the management of high-level disposal systems were proposed to select sites for deep geological disposal facilities and to implement business strategies. Leading countries such as Finland, Sweden and France have developed and applied safety cases to verify the safety of deep geological disposal systems. By examining the regulatory status of foreign leading countries, we analyze the safety cases ranging from the site selection stage of the deep geological disposal system to the securing of the permanent disposal system to the investigation, analysis, evaluation, design, construction, operation, and closure. Based on this analysis, we will develop safety case elements for long-term safety of deep geological disposal systems suitable for domestic situation. To systemically analyze data based on safety cases, we have established a database of deep geological disposal system regulations in leading foreign countries. Artificial intelligence text mining and data visualization techniques are used to provide database in dashboard form rather than simple lists of data items, which is a limitation of existing methods. This allows regulatory developers to understand information more quickly and intuitively and provide a convenient interface so that anyone can easily access the analyzed data and create meaningful information. Furthermore, based on the accumulated bigdata, the artificial intelligence learns and analyzes the information in the database through deep learning, and aims to derive a more accurate safety case. Based on these technologies, this study analyzed the legal systems, regulatory standards, and cases of major international leading countries and international organizations such as the United States, Sweden, Finland, Canada, Switzerland, and the IAEA to establish a database management system. To establish a safety regulation base suitable for the domestic deep geological disposal environment, the database is provided as data to refer to and apply systematic information management on regulatory standards and regulatory cases of overseas leading countries, and it is expected that it will play a key role as a forum for understanding and discussing the level of safety of deep geological disposal system among stakeholders.

송충살이고치벌아과의 Aleiodes의 미기록종 4종(A. lipwigi, A. guidaae, A. subemarginatus, A. caprinus)을 국내 최초로 보고한다. 본 연구에 선 종의 동정을 위한 진단, 분포, 삽화 및 DNA 바코드 서열을 제공한다.

Metal-additive manufacturing techniques, such as selective laser sintering (SLS), are increasingly utilized for new biomaterials, such as cobalt-chrome (Co-Cr). In this study, Co-Cr gas-atomized powders are used as charge materials for the SLS process. The aim is to understand the consolidation of Co-Cr alloy powder and characterization of samples sintered using SLS under various conditions. The results clearly suggest that besides the matrix phase, the second phase, which is attributed to pores and oxidation particles, is observed in the sintered specimens. The as-built samples exhibit completely different microstructural features compared with the casting or wrought products reported in the literature. The microstructure reveals melt pools, which represent the characteristics of the scanning direction, in particular, or of the SLS conditions, in general. It also exposes extremely fine grain sizes inside the melt pools, resulting in an enhancement in the hardness of the as-built products. Thus, the hardness values of the samples prepared by SLS under all parameter conditions used in this study are evidently higher than those of the casting products.

High-entropy alloys (HEAs) are attracting attention because of their excellent properties and functions; however, they are relatively expensive compared with commercial alloys. Therefore, various efforts have been made to reduce the cost of raw materials. In this study, MIM is attempted using coarse equiatomic CoCrFeMnNi HEA powders. The mixing ratio (powder:binder) for HEA feedstock preparation is explored using torque rheometer. The block-shaped green parts are fabricated through a metal injection molding process using feedstock. The thermal debinding conditions are explored by thermogravimetric analysis, and solvent and thermal debinding are performed. It is densified under various sintering conditions considering the melting point of the HEA. The final product, which contains a small amount of non-FCC phase, is manufactured at a sintering temperature of 1250oC.

Subunit vaccines are being developed as a potential therapy for preventing microbial pathogen infection. In this study, the immunogenicity of recombinant Brucella (B.) abortus Fe/Mn superoxide dismutase (rFe/Mn SOD) protein as a subunit vaccine against B. abortus was investigated in BALB/c mice model. Brucella Fe/Mn SOD gene was cloned into a pcold-TF DNA vector. The bacterial recombinant protein was expressed using the Escherichia coli DH5α strain with a size of 82.50 kDa. The western blotting assay showed that rFe/Mn SOD reacted with Brucella-positive serum, indicating the potential immunoreactivity of this recombinant protein. After the second and third vaccinations, the peripheral CD4+ T cell population was increased significantly in the rFe/Mn SOD-immunized mice group compared to the PBS control group. Moreover, immunization of this recombinant protein increased the CD4+ T cell population from the first vaccination to the third vaccination. Meanwhile, the CD8+ T cells were slightly enhanced after the second vaccination compared to the first vaccination and compared to control groups. Fourteen days after the bacterial infection, the splenomegaly and the number of bacteria in the spleen were evaluated. The result showed that both rFe/Mn SOD and positive control RB51 decreased the bacterial replication in the spleen and the splenomegaly compared to control groups. Altogether, these results suggested that rFe/Mn SOD could induce host immunity against B. abortus infection.

Over the past decade, platelet-rich plasma (PRP) and platelet derivatives have been widely investigated in the field of regenerative medicine due to their high concentrations of platelet- related growth factors, cytokines, and other proteins. Recently, many clinical studies have suggested their regenerative therapeutic efficacy in treating several disorders in medical field. However, their therapeutic applications are not well characterized in veterinary medicine as in human and experimental animals. This article reviews functional roles of platelets, scientific concepts, and clinical use of PRP and platelet derivatives in veterinary medicine. It also presents guidelines for veterinary applications of PRP in the future.

PURPOSES : The aim of this study is to evaluate the stripping resistance of a bead coating via the Hamburg wheel tracking test and image analysis. METHODS : First, the stripping resistance of the bead coating was evaluated via the Hamburg wheel tracking test. A pneumatic wheel with a load of 175±2 N was used to simulate repeated skid cycles. Several bead coating mixtures with different numbers of coating layers, i.e., zero, one, two, three, and four layers, i.e., zero, one, two, three, and four layers,were conducted. Finally, an image analysis program was developed to analyze surface images captured from the Hamburg wheel tracking test. RESULTS : The results show that the samples with more coating layers exhibit higher stripping resistance. After 500 stripping cycles, the percentage of bead loss is 4% to 28%. At 80% bead loss, the mixture with one coating layer presents more skid cycles than the control sample without a coating layer. CONCLUSIONS : Incorporating a coating layer can improve the stripping resistance of glass beads under repeated skid cycles. Additionally, an image analysis program is established in this study to determine the percentage of bead loss caused by the stripping test.

Kale (Brassica oleracea var. acephala) is one of the most frequently consumed leafy vegetables globally, as it contains numerous nutrients; essential amino acids, phenolics, vitamins, and minerals, and is particularly rich in glucosinolates. However, the differences in the biosynthesis of glucosinolates and related gene expression among kale cultivars has been poorly reported. In this study, we investigated glucosinolates profile and content in three different kale cultivars, including green (‘Man-Choo’ and ‘Mat-Jjang’) and red kale (‘Red-Curled’) cultivars grown in a vertical farm, using transcriptomic and metabolomic analyses. The growth and development of the green kale cultivars were higher than those of the red kale cultivar at 6 weeks after cultivation. High-performance liquid chromatography (HPLC) analysis revealed five glucosinolates in the ‘Man-Choo’ cultivar, and four glucosinolates in the ‘Mat-Jjang’ and ‘Red-Curled’ cultivars. Glucobrassicin was the most predominant glucosinolate followed by gluconastrutiin in all the cultivars. In contrast, other glucosinolates were highly dependent to the genotypes. The highest total glucosinolates was found in the ‘Red-Curled’ cultivar, which followed by ‘Man-Choo’ and ‘Mat-Jjang’. Based on transcriptome analysis, eight genes were involved in glucosinolate biosynthesis. The overall results suggest that the glucosinolate content and accumulation patterns differ according to the kale cultivar and differential expression of glucosinolate biosynthetic genes.