• Journal of Intelligence and Information Systems
• /
• v.29 no.1
• /
• pp.249-263
• /
• 2023

Collaborative Filtering, a representative recommendation system methodology, consists of two approaches: neighbor methods and latent factor models. Among these, the latent factor model using matrix factorization decomposes the user-item interaction matrix into two lower-dimensional rectangular matrices, predicting the item's rating through the product of these matrices. Due to the factor vectors inferred from rating patterns capturing user and item characteristics, this method is superior in scalability, accuracy, and flexibility compared to neighbor-based methods. However, it has a fundamental drawback: the need to reflect the diversity of preferences of different individuals for items with no ratings. This limitation leads to repetitive and inaccurate recommendations. The Adaptive Deep Latent Factor Model (ADLFM) was developed to address this issue. This model adaptively learns the preferences for each item by using the item description, which provides a detailed summary and explanation of the item. ADLFM takes in item description as input, calculates latent vectors of the user and item, and presents a method that can reflect personal diversity using an attention score. However, due to the requirement of a dataset that includes item descriptions, the domain that can apply ADLFM is limited, resulting in generalization limitations. This study proposes a Generalized Adaptive Deep Latent Factor Recommendation Model, G-ADLFRM, to improve the limitations of ADLFM. Firstly, we use item ID, commonly used in recommendation systems, as input instead of the item description. Additionally, we apply improved deep learning model structures such as Self-Attention, Multi-head Attention, and Multi-Conv1D. We conducted experiments on various datasets with input and model structure changes. The results showed that when only the input was changed, MAE increased slightly compared to ADLFM due to accompanying information loss, resulting in decreased recommendation performance. However, the average learning speed per epoch significantly improved as the amount of information to be processed decreased. When both the input and the model structure were changed, the best-performing Multi-Conv1d structure showed similar performance to ADLFM, sufficiently counteracting the information loss caused by the input change. We conclude that G-ADLFRM is a new, lightweight, and generalizable model that maintains the performance of the existing ADLFM while enabling fast learning and inference.

• Journal of Intelligence and Information Systems
• /
• v.29 no.1
• /
• pp.79-105
• /
• 2023

Due to unprecedented extreme weather events such as global warming and climate change, many parts of the world suffer from severe pain, and economic losses are also snowballing. In order to address these problems, 'The Paris Agreement' was signed in 2016, and an intergovernmental consultative body was formed to keep the average temperature rise of the Earth below 1.5℃. Korea also declared 'Carbon Neutrality in 2050' to prevent climate catastrophe. In particular, it was found that the increase in temperature caused by greenhouse gas emissions hurts the environment and society as a whole, as well as the export-dependent economy of Korea. In addition, as the diversification of transportation types is accelerating, the change in means of choice is also increasing. As the development paradigm in the low-growth era changes to urban regeneration, interest in idle railway sites is rising due to reduced demand for routes, improvement of alignment, and relocation of urban railways. Meanwhile, it is possible to partially achieve the solar power generation goal of 'Renewable Energy 3020' by utilizing already developed but idle railway sites and take advantage of being free from environmental damage and resident acceptance issues surrounding the location; but the actual use and plan for these solar power facilities are still lacking. Therefore, in this study, using the big data provided by the Korea National Railway and the Renewable Energy Cloud Platform, we develop an algorithm to discover and analyze suitable idle sites where solar power generation facilities can be installed and identify potentially applicable areas considering conditions desired by users. By searching and deriving these idle but relevant sites, it is intended to devise a plan to save enormous costs for facilities or expansion in the early stages of development. This study uses various cluster analyses to develop an optimal algorithm that can derive solar power plant locations on idle railway sites and, as a result, suggests 202 'actively recommended areas.' These results would help decision-makers make rational decisions from the viewpoint of simultaneously considering the economy and the environment.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.21 no.6
• /
• pp.21-35
• /
• 2022

With the support of local and central governments, various incentive policies for "green" cars have been established, and the number of electric vehicle users has been rapidly increasing in recent years. As a result, much attention is being given to establishing a user-centered charging infrastructure. A standard for the number of electric vehicle chargers to be supplied is being prepared based on building characteristics, but there is quite limited research on the appropriate ratio of slow and fast chargers based on the characteristics of each place. Therefore, this study derived an appropriate penetration ratio based on data about the distribution ratio of common slow chargers. These data were collected using a survey of actual electric vehicle users. Next, an analysis was done on how to categorize the needs of charging environments and to determine what criteria or characteristics to use for categorization. Based on the results of the survey analysis, three types of places were derived. Type-1 places require 10% of chargers to be slow chargers, Type-2 places require 40-60% of chargers to be slow chargers (i.e., around equal distribution of slow and fast chargers), and Type-3 places require more than 80% of chargers to be slow chargers. The required levels of slow chargers were classified by place type and by individual using latent class cluster analysis, which made it possible to categorize them into five clusters related to socioeconomic variables, vehicle characteristics, traffic, and charging behaviors. It was found that there was a high correlation between charging behavior, weekend travel behavior, gender, and income. The results and insights from this study could be used to establish charging infrastructure policies in the future and to prepare standards for supplying charging infrastructure according to changes in the electric vehicle market.

• Korean Journal of Plant Resources
• /
• v.36 no.1
• /
• pp.62-90
• /
• 2023

This study was conducted to study the vascular plants of Balbaeksan, located on the border between Gyeongsan-si and Cheongdo-gun, Gyeongsangbuk-do. It was performed 13 times from September 2020 to November 2021, and the investigated vascular plants were identified as a total of 93 families, 302 genera, 485 species, 10 subspecies, 40 varieties, 13 varieties, and 545 taxa. Endemic plants were investigated in 18 taxa, 9 taxa of rare plants; Vulnerable (VU) 5 taxa, Least concerned (LC) 3 taxa, Data Deficient (DD) 1 taxa, Red List plants 5 taxa; 3 taxa of least concerned (LC), 2 taxa of Not Evaluated (NE), 183 taxa of biological resources subject to export approval, 78 taxa of specific taxonomy; There were 9 taxa of grade IV, 20 taxa of grade III, 20 taxa of grade II, and 28 taxa of grade I. 4 taxa of plants subject to climate change adaptation; 2 taxa of endemic plants, 1 taxa of southern plants, 1 taxa of northern plants, 4 taxa of climate change biomarker species, 2 taxa of candidate species of climate change biomarker, 65 taxa of exotic plants, and 3 taxa of ecosystem disturbance species. Plants 235 taxa (43.1%), medicinal plants 257 taxa (47.2%), ornamental plants 148 taxa (27.2%), woody plants 54 taxa (9.9%), herbaceous plants 37 taxa (6.8%), industrial plants 15 Taxa (2.8%), 11 taxa (2.0%) of fiber-bearing plants, and plants of unknown use were identified as 73 taxa (13.4%). The flora of Mt. Balbaek was investigated for the first time, and plants to be preserved in order to secure useful plant resources and biological sovereignty were identified. So continuous monitoring is necessary to prevent the introduction of foreign plants that may be caused by the recent increase in the number of people coming and going.

• Journal of Life Science
• /
• v.33 no.5
• /
• pp.406-413
• /
• 2023

In this study, the various physiological and anti-inflammatory activities of Sanguisorba officinalis L. roots (SR) were assessed for potential use as functional cosmetic materials. As a result of measuring electron-donating abilities to determine the antioxidant ability of SR extract, activity increased as the concentration increased, showing an excellent antioxidant capacity of 93.8% at a 1,000 ㎍/ml concentration. Further, the antioxidant power of SR extract, which was determined using an ABTS⁺ assay measurement, was more than 99% at concentrations of 50 ㎍/ml or more, while the tyrosinase inhibition rate was 37.7% at the highest concentration of 1,000 ㎍/ml. Consequently, the elastase and collagenase inhibition of SR extract measured 84.9% and 90.3%, respectively, at a 1,000-㎍/ml concentration. As a result of confirming the survival rate of Raw 264.7 cells, the cell survival rate was determined to be 80% or more below a 100 ㎍/ml concentration, and subsequent cell-related experiments were conducted at concentrations below 100 ㎍/ml. Furthermore, after applying a NO assay to identify anti-inflammatory activity, it was confirmed that SR extract had an inhibitory rate of 50.8% at a concentration of 500 ㎍/ml, and it was excellent at suppressing the inflammatory expression. As a result of verifying protein expression by treating SR extract in Raw 264.7 cells, it was confirmed that expression was inhibited concentrated in all factors. Therefore, it is judged that SR can be used as a functional cosmetic material with antioxidant, whitening, and wrinkle-improving physiological effects and anti-inflammatory activities.

• Korean Journal of Agricultural and Forest Meteorology
• /
• v.25 no.1
• /
• pp.28-36
• /
• 2023

Solar radiation that is measured at relatively small number of weather stations is one of key inputs to crop models for estimation of crop productivity. Solar radiation products derived from GK-2A and Himawari 8 satellite data have become available, which would allow for preparation of input data to crop models, especially for assessment of crop productivity under an agrivoltaic system where crop and power can be produced at the same time. The objective of this study was to compare the degree of agreement between the solar radiation products obtained from those satellite data. The sub hourly products for solar radiation were collected to prepare their daily summary for the period from May to October in 2020 during which both satellite products for solar radiation were available. Root mean square error (RMSE) and its normalized error (NRMSE) were determined for daily sum of solar radiation. The cumulative values of solar radiation for the study period were also compared to represent the impact of the errors for those products on crop growth simulations. It was found that the data product from the Himawari 8 satellite tended to have smaller values of RMSE and NRMSE than that from the GK-2A satellite. The Himawari 8 satellite product had smaller errors at a large number of weather stations when the cumulative solar radiation was compared with the measurements. This suggests that the use of Himawari 8 satellite products would cause less uncertainty than that of GK2-A products for estimation of crop yield. This merits further studies to apply the Himawari 8 satellites to estimation of solar power generation as well as crop yield under an agrivoltaic system.

• Journal of the Korean Society of Earth Science Education
• /
• v.16 no.1
• /
• pp.143-152
• /
• 2023

This study was conducted on 115 students from 4 classes of 2nd year intensive courses at B University of Education. Elementary preservice teachers were surveyed on what teaching methods they were aware of regarding the subject of elementary science astronomy. Recognition data of 80 people from whom the questionnaire was collected were analyzed. For this study, a 5-day survey was conducted. Elementary preservice teachers complained of difficulties in teaching methods in the astronomy area among various areas of the science department. The purpose of this study was to find out what difficulties elementary preservice teachers have in teaching elementary science astronomy topics, and to find more efficient teaching methods for teaching astronomy topics. The topic of the survey was set by discussing with the preparatory elementary teachers about what kind of survey to use in teaching the subject of elementary science astronomy. There are many topics for elementary science astronomy, but two questionnaires were prepared focusing on the unit on the earth and the moon. 'What does the earth look like?' in Unit 4 (1/10) of the 3rd year, 1st semester In Unit 2 (1/11) of the 1st semester of the 6th grade, it was set as 'What does the moon look like?'. Candidly describe how to teach the subject of astronomy to elementary school students by mobilizing all the background knowledge of preparatory elementary teachers. The results of these surveys were visualized and displayed using Netminer as a language analysis method, and the contents of the responses to the actual surveys by pre-service elementary school teachers were described and interpreted. Based on these results, preparatory elementary teachers tried to suggest a more efficient teaching method for the subject of elementary science astronomy. In addition, basic procedures and methods for lecturing on the subject of elementary science astronomy were presented. A more efficient teaching method for teaching elementary science astronomy subjects to pre-service elementary teachers was suggested.

• Journal of Environmental Health Sciences
• /
• v.49 no.2
• /
• pp.66-77
• /
• 2023

Background: Along with the increase in consumer interest in and consumption of tattoo products, the controversy over harmful heavy metals associated with the use of tattoo cosmetics is also increasing. Therefore, investigation of hazardous metals in these tattoo products is needed. Objectives: This study was performed to provide useful data for establishing reasonable standards to securely manage tattoo cosmetics, tattoo stickers, and tattoo inks distributed in the market. Methods: Thirteen kinds of hazardous metal contents (Pb, As, Cd, Sb, Ni, Co, Cu, Cr, Se, Ba, Zn, Sn, and Hg) were analyzed for 23 tattoo cosmetics, ten tattoo stickers, and 16 tattoo inks. Hg was measured through the combustion-gold amalgamation method, and other hazardous metals were measured by inductively coupled plasma-mass spectrometry (ICP-MS) after acidic decomposition using a microwave apparatus. Results: The detected ranges of Pb, As, Cd, Sb, Ni, and Hg in tattoo cosmetics were 0.07~1.18, 0.06~0.41, ND~0.07, 0.01~3.44, 0.12~2.75, and ND~0.01 ㎍/g, respectively. All of the hazardous metals detected were below the recommended maximum standards of the Ministry of Food and Drug Safety. The mean amount of Pb detected in tattoo stickers for children was 0.24 ㎍/kg and Cd was not detected, meaning both metals met the recommended criteria. There was no statistically significant difference in all measured metals between children's tattoo stickers and adults' tattoo stickers. In the results of the study on the hazardous metal content of tattoo inks, four products (25%) for Pb, one product (6%) for As, 13 products (81%) for Ni, four products (25%) for Cu, and five products (31%) for Zn exceeded the recommended standards approved by the government. The highest predicted exposure amount for hazardous metals exceeding the recommended level of tattoo inks in a single tattooing was 5.69 ㎍/kg for Ni, 8.51 ㎍/kg for Zn, 0.44 ㎍/kg for Pb, 8.07 ㎍/kg for Cu, 0.44 ㎍/kg for As, and 71.36 ㎍/kg for Ba. Conclusions: It is necessary to prepare criteria for content limitation for the management of Co, Cr, Ba and Se tattoo cosmetics, and tattoo inks require thorough quality control.

• Clean Technology
• /
• v.29 no.3
• /
• pp.193-199
• /
• 2023

Biomass has been used to secure renewable energy certificates (REC) in domestic and overseas coal-fired power plants. In recent years, biofuel has been diversified from traditional wood pellets to non-woody biomass. Non-woody biomass has a higher content of alkaline metals such as K and Na than wood-based biomass, resulting in a lower melting point and an increase in slagging on boiler tubes, which reduces boiler efficiency. This study analyzed the effect of kaolin, an additive commonly used to increase melting points, on biomass co-firing to coal through thermochemical equilibrium calculations. In a previous experiment on biomass co-firing to coal conducted at 80 kW_th, it was interpreted that the use of kaolin actually increased the amount of fouling. In this study, analysis showed that when kaolin was added, aluminosilicate compounds were generated due to Al₂O₃, which is abundant in coal, and mullite was formed. Thus, it was confirmed that the amount of slag increased when more kaolin was used. Further analysis was conducted by increasing the biomass co-firing rate from 0% to 100% at 10% intervals, and the results showed non-linear liquid slag generation. As a result, it was found that the least amount of liquid slag was generated when the biomass co-firing rate was between 50 and 60%. The phase diagram analysis showed that high melting point compounds such as leucite and feldspar were most abundantly generated under these conditions.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2023.05a
• /
• pp.67-67
• /
• 2023

수문관측의 핵심은 강우-유출 관계다. 하천유량 생산을 위해 수많은 지점에서 유량측정을 수행한다. 그러나 유량자료의 신뢰도는 높지 않다. 그리고 댐 유입량 산정에 아무 도움이 되지 않고 있다. 더구나 저수지의 경우는 유입량 자료도 없이 운영되고 있다. 저수지, 댐에 물이 고여 있는데 이를 활용하면 유입량을 계산할 수 있고, 그 지점에서 유량이 얼마인지 고품질로 생산할 수 있다. 여기서는 총저수량 260만m³, 유역면적 3.7km²인 감포댐에 적용하여 저수량 자료를 활용하여 유입량의 신뢰도를 얼마나 개선시킬 수 있는지 분석한 결과는 다음과 같다. 여기서 적용 기간은2020.9.1.~9.14., 2022.9.5.~9.6 등 2개 사상이고, ONE 모형에 의해 10분 단위로 유출량을 모의했다. 모의 방법은 총유량을 같게 하는 방법과 저수위 오차를 최소로 하는 방법 등 두 가지로 했다. 첫째, 2020.9.1.~9.14. 사상은 강우량은 10분 최대 19.0mm, 총 127.0mm였다. 총유량을 같게 하는 경우 유입량은 10분 최대 5.4m³/s, 총 40만m³로 모의돼, 유출률 85.5%로 나타났고, 관측은 10분최대 4.7m³/s, 총 40만m³, 유출률 85.3%로 나타났다. 유량 신뢰도는 RMSE 0.491mm, NSE 0.237, R²는 0.455로 나타났다. 이 경우 저수위 모의 신뢰도는 RMSE 0.600m, NSE 0.158, R²는 0.893로나타났다. 저수량 오차를 최소로 한 경우 유입량은 10분 최대 4.0m³/s, 총 28만m³로 모의돼, 유출률 59.4%로 나타났고, 관측은 10분 최대 4.0m³/s, 총 28만m³, 유출률 85.3%로 나타났다. 유량 신뢰도는 RMSE 0.425mm, NSE 0.430, R²는 0.507로 나타났다. 이 경우 저수위 모의 신뢰도는 RMSE 0.110m, NSE 0.972, R²는 0.995로 높았다. 둘째, 2022.9.5.~9.6. 사상은 강우량은 10분 최대 32.3mm, 총 196.0mm였다. 총유량을 같게 하는 경우 유입량은 10분 최대 64.5m³/s, 총 59만m³로 모의돼, 유출률 81.6%로 나타났고, 관측은 10분 최대 80.1m³/s, 총 59만m³, 유출률 81.6%로 나타났다. 유량 신뢰도는 RMSE 1.832mm, NSE 0.960, R²는 0.984로 나타났다. 이 경우 저수위 모의 신뢰도는 RMSE 0.323m, NSE 0.968, R²는 0.999로나타났다. 저수량 오차를 최소로 한 경우 유입량은 10분 최대 80.1m³/s, 총 66만m³로 모의돼, 유출률 91.6%로 나타났고, 관측은 10분 최대 80.1m³/s, 총 59만m³, 유출률 81.8%로 나타났다. 유량 신뢰도는 RMSE 2.120mm, NSE 0.947, R2는 0.949로 나타났다. 이 경우 저수위 모의 신뢰도는 RMSE 0.153m, NSE 0.993, R²는 0.997로 높았다. 종합하면 저수량 오차가 최소가 되도록 하천 유출량을 모의하면 결과적으로 하천유량의 신뢰도를 향상시키는 것이라 말할 수 있다.