Earticle

With the recent introduction of artificial intelligence (AI) technology, the use of data is rapidly increasing, and newly generated data is also rapidly increasing. In order to obtain the results to be analyzed based on these data, the first thing to do is to classify the data well. However, when classifying data, if only one classification technique belonging to the machine learning technique is applied to classify and analyze it, an error of overfitting can be accompanied. In order to reduce or minimize the problems caused by misclassification of the classification system such as overfitting, it is necessary to derive an optimal classification by comparing the results of each classification by applying several classification techniques. If you try to interpret the data with only one classification technique, you will have poor reasoning and poor predictions of results. This study seeks to find a method for optimally classifying data by looking at data from various perspectives and applying various classification techniques such as LDA and QDA, such as linear or nonlinear classification, as a process before data analysis in data analysis. In order to obtain the reliability and sophistication of statistics as a result of big data analysis, it is necessary to analyze the meaning of each variable and the correlation between the variables. If the data is classified differently from the hypothesis test from the beginning, even if the analysis is performed well, unreliable results will be obtained. In other words, prior to big data analysis, it is necessary to ensure that data is well classified to suit the purpose of analysis. This is a process that must be performed before reaching the result by analyzing the data, and it may be a method of optimal data classification.

연구목적: 아차사고의 분석을 통하여 재해를 예측하여 사고를 예방하는 목적이 있다. 연구방법: 본 연구에 서는 건설 현장의 아차사고 문헌조사 및 데이터를 수집하고, 설문조사를 실시하여 아차사고 연결 가능성 분류를 위해 로지스틱 회귀분석 및 의사결정나무 분석을 이용하였다. 연구결과: 아차사고 types이 정신적, 신체적, 안전습관 행동에 미치는 영향을 분석한 결과, 신체에 영향력이 높은 요인은 아차사고 관리의 필요 성, 직종은 전기·정보통신, 건강 상태 순으로, 정신적 요인에서 공사 규모가 영향력이 높았으며, 경험 공종, 중상자 수, 직종 순으로 습관 행동 요인에 영향력이 높은 요인은 착각, 부적절한 작업지시, 신체 부위 순으 로 분석되었다. 의사결정나무 분석을 통해 아차사고가 놀랄 정도의 사고로 연결가능성에 영향을 미치는 요인과 패턴을 확인하였다. 결론: 건설현장관계자는 아차사고 관찰을 고려하여 정신적. 신체적 측면의 아 차사고와의 연관성에 대한 구체적 관리가 실행되어야 하며, 부적절한 작업지시가 아차사고로 연결되지 않 도록 인원 배치, 작업계획, 작업절차 및 방법, 피드백을 통해 중대재해가 저감하는 작업환경을 기대한다.

Purpose: The purpose is to prevent accidents by predicting disasters through the analysis of near-miss. Method: In this study, a near-miss literature review and data were collected at construction sites, and a questionnaire survey was conducted to use logistic regression analysis and decision tree analysis to classify the possibility of near-miss connection. Result: As a result of analyzing the effects of near-miss types on mental, physical, and safety habits and behaviors, the factor with a high influence on the body is the need for near-miss management, the type of job is electricity·information communication, and health status in order, and the mental factor is the construction scale The influence was high, and the factors with the highest influence on the habit behavior factors were analyzed in the order of experience, number of serious injuries, and occupation in order of illusion, inappropriate work instructions, and body parts. Through decision tree analysis, factors and patterns that affect the possibility of a near-miss being a surprise accident were identified. Conclusion: Construction site officials consider the observation of near-miss and mentally and physically. Specific management of the relevance of physical aspects to near-miss should be implemented, and a work environment in which serious accidents are reduced is expected through personnel allocation, work plans, work procedures and methods, and feedback so that inappropriate work instructions do not lead to near-miss.

오늘날 도시관리의 정책의 흐름은 과거 성장중심에서 관리‧보존이 공존하는 도시재생 중심으로 무게중심이 전환되어 왔으며, `19년 현재 정부지원 도시재생뉴딜사업으로 전국에 총 329개소와 서울시에서 47개소의 도시재생활성화지역이 지정‧운영되고 있다. 따라서 본 연구는 법 도입초기 진행된 도시재생사업의 공공예산사업이 끝나는 시점에서 도시재생 유형별 도시재생사업이 어떠한 특성으로 구성되었는지에 대하여 고시된 활성화계획의 사업구성 내용을 토대로 유형화하고 도시재생사업의 본래의 취지에 부합하게 주민이 원하는 사업이 도시재생의 기본목표와 정합성을 가지는지 활성화계획의 계획적 적합성을 실증분석 하였다. 그 결과 도시재생 유형별 목적에 부합하게 근린재생형은 주거/생활환경재생을 위한 물리적 사업의 비중이 높고, 중심시가지형은 지역자산에 기반한 산업‧역사문화 특화를 위한 프로그램 사업이 높음을 알 수 있으며, 경제기반형은 새로운 경쟁력을 창출을 위해 산업, 기반시설 등의 물리적 사업이 중심이 되어 계획된 것으로 분석된다. 또한 도시재생지역의 주민(상인) 설문조사를 통한 분석결과 활성화계획이 주민의 요구에 부합하게 수립되었음을 검증 할 수 있었다. 이 연구는 유형별 도시재생사업을 모니터링하고 성과를 검증 할 때도 이와 같은 도시재생사업 유형화에 따라 실제적인 효과가 측정될 수 있게 평가지표 구성에 기초자료로 활용 될 수 있을 것으로 기대된다.

Nowadays, the flow of urban management policy has shifted from development-focused one to urban regeneration oriented one where maintenance and preservation coexist. Therefore, this study categorized detailed projects based on the project composition contents of the announced revitalization plan and conducted empirical analysis on appropriateness as a plan whether projects desired by residents are in keeping with the basic goal of urban regeneration, that is the true purpose of urban regeneration business. As a result and according to the purpose of urban regeneration by type, it is analyzed that Neighborhood Regeneration has a high proportion of physical projects for residential / living environment regeneration, Utilization of local asset Regeneration has a high number of programs specializing in industrial and historical culture based on local assets, and Urban Economic Regeneration has focused on physical projects such as industries and infrastructure as the center to create new competitiveness. In addition, according to the result of surveys conducted by residents, urban regeneration revitalization plan has been established in consistent with goals of urban regeneration by type and demands from residents. This study is expected to be used as basic data in the construction of evaluation indicators so that actual effects can be measured according to the type of urban regeneration project even when monitoring urban regeneration projects by type and verifying their accomplishments.

Small and medium-sized cities charge an important role in provincial areas which help to strengthen the competitiveness. In spite of efforts to keeping the balance between Seoul Metropolitan Area(SMA) and provinces, there has not yet been any significant effects. Because, there are lots of problems such as insufficient consideration of local conditions, uniform contents of projects and complaints of regions. Therefore, improvement of current systems is significant object to vitalize small and medium-sized cities. This study focused on the comparison of the small and medium-sized cities located in SMA and provinces. And we classified those cities through cluster analysis and drew up a plan for improvement of current system. The findings from analyses were as follows : First, the unbalance between SMA and provinces were getting worse in the demographic and financial aspects. Especially, the provincial small and medium-sized cities which have population less than 300,000 were faced with serious decrease of the population and an aging society. Second, cluster analysis of those cities revealed that most of small and medium-sized cities were fallen into such groups with shortage of the infrastructure, low economic self-sufficiency and weak infrastructure to growth. Finally, regional development policies include dissatisfaction to solve the fundamental problem of small and medium-sized cities due to the lack of linkage between policies decentralized public investment, and so on. The findings in this study suggests the need for integration of regional development policies and improvement of the current support system to vitalize small and medium-sized cities.

Class imbalance in datasets presents a significant challenge in machine learning, often causing traditional classification algorithms to exhibit bias towards majority classes while underrepresenting minority classes, which may be of crucial importance in various applications. This classification challenge is further exacerbated by the presence of label noise, which impedes the identification of optimal decision boundaries between classes and potentially leads to model overfitting. While extensive research has addressed class imbalance and label noise as separate phenomena, there remains a notable gap in the literature regarding their concurrent occurrence in datasets, specifically in the domain of imbalanced classification with label noise (ICLN). This review aims to bridge this gap by conducting an extensive analysis of existing methodologies addressing ICLN challenges. Our review encompasses approaches across diverse categories, including resampling techniques, ensemble methods, cost-sensitive learning, deep learning, active learning, meta-learning, and hybrid methodologies. Through rigorous empirical evaluation, we compare representative methods from each category using synthetic and real-world datasets, revealing a trade-off between minority class preservation, noise robustness, and computational efficiency. Our findings reveal that algorithm effectiveness is fundamentally dataset-dependent, with deep learning methods excelling on complex datasets while resampling approaches achieve competitive performance with lower computational cost. Statistical significance analysis validates our empirical observations, and we identify concrete future research directions for advancing ICLN methodologies.

This study utilized a standardization and cluster analysis technique for the selection and classification of beneficial bacteria. A set of synthetic data consisting of 100 individual variables with three characteristics was created for analysis. The three characteristics assigned to each independent variable were designated to have different numeric scales, averages, and standard deviations. The variables were bacterial isolates at random, and the three characteristics were fermentation products, including cell yield, antioxidant activity of culture, and enzyme production. A standardization method utilizing a standard normal distribution equation to record fermentation yields of each isolate was employed to weight their different numeric scales and deviations. Following transformation, the data set was analyzed by cluster analysis. The Manhattan method for dissimilarity matrix construction along with complete linkage technique, an agglomerative method for hierarchical cluster analysis, was employed using statistical computing program R. A total of 100 isolates were classified into groups A, B, and C. In a comparison of the characteristics of each group, all characteristics in groups A and C were higher than those of group B. Isolates displaying higher cell yield were classified as group A, whereas those isolates showing high antioxidant activity and enzyme production were assigned to group C. The results of the cluster analysis can be useful for the classification of numerous isolates and the preparation of an isolation pool using numerical or statistical tools. The present study suggests that a simple technique can be applied to screen and select beneficial microbes using the freely downloadable statistical computing program R.

PurposeThis study aimed to classify self-care patterns among Korean adults with prediabetes using an unsupervised machine learning approach. The classification was grounded in Orem’s Self-Care Theory, focusing on self-care demands, self-care agencies, and self-care behaviors. MethodsA secondary data analysis was conducted using the 2023 Korea National Health and Nutrition Examination Survey. Variables were selected and categorized according to the theoretical components of Orem’s model. Principal component analysis was applied for dimensionality reduction, followed by K-means clustering to identify distinct self-care pattern groups. All variables were standardized using min-max normalization. Group differences were examined using analysis of variance and the chi-square test. ResultsThree self-care pattern groups were identified: the high self-care performance group, the latent self-care risk group, and the self-care vulnerable group. These groups exhibited distinct profiles across self-care demands, agencies, and behaviors. Significant intergroup differences were also observed in education level, income, health literacy, fasting blood glucose, and hemoglobin A1c levels. ConclusionSelf-care patterns among adults with prediabetes can be effectively classified through unsupervised learning techniques. The findings highlight the importance of developing tailored nursing interventions that consider multidimensional self-care profiles. This study underscores the applicability of Orem’s Self-Care Theory and demonstrates the potential of machine learning in identifying at-risk subgroups for early intervention.

The purpose of this study was to verify classification performance and the difference analysis between gender using optimal warping paths of dynamic time warping (DTW) and to examine the usefulness of root mean square error (RMSE) represented by the perpendicular distance from the optimal warping path to the diagonal. A 3-dimensional motion analysis experiment was performed with 24 healthy adults (male=12, female=12) in their 20s of age without gait-related diseases or injuries for the past 6 months to collect gait data. This study performed a DTW 132 times in total (male=62, female=62) for the flexion angle of the right leg’s hip, knee, and ankle joints. Then, the global cost and the RMSE of the optimal warping paths were calculated and normalized. The differ-ence analysis was performed by independent t-test. Machine learning was performed to test the classification performance using the neural network, support vector machine, and logistic regression model among the supervised models. Results analyzed using global cost and RMSE for hip, knee, and ankle joints showed a statistically significant differ-ence between genders in global cost and RMSE for hip and knee joints but not for ankle joints using RMSE. Considering both area under the receiver operating characteristic curve and F1-score, the logistic re-gression model has been evaluated as the most suitable for gender classification using the global cost or RMSE. This study demonstrated that optimal warping paths could be used for statistical difference anal-ysis and classification analysis.

국내 저수지 제체의 안전도는 안전점검을 통한 안전등급으로 판정된다. 저수지 안전등급 제도는 2009년부터 시행되어 왔으나, 안전등급에 대한 정량적 안전 성능 연구는 미진한 실정이다. 2020년에 발생한 25건의 저수지 제체 붕괴사고는 안전등급에 대한 정량적 안전성능 연구에 소중한 자료가 된다. 본 연구에서는 전국 15,732개소 저수지의 안전등급과 강우 자료를 확보하여 강우량이 유사한 강우 그룹과 안전등급이 동일한 안전등급 그룹에서의 통계적 제체 붕괴확률을 추정하였다. 2020년 의 경우 B 안전등급에서 12개소, C 안전등급에서 10개소, D 안전등급에서 3개소의 제체 붕괴사고가 있었다. 제체 붕괴사고가 전무한 A 안전등급 그룹은 다른 안전등급 그룹에 비해 매우 탁월한 안전성 능을 보였으며, B와 C 안전등급 그룹은 전반적으로 비슷한 안전성능을 보였다. 극한 기상 조건에서 안전등급에 따른 안전성능 차이가 큰 것으로 나타났다.

The reservoir’s safety rating means its safety level, which is evaluated by the safety inspection on its embankment and accessory structures in Korea. Although the safety rating system have been enforced from 2009, quantitative studies on the safety performance of the safety rating are still poor. 25 embankment breaches in 2020 are valuable to the quantitative studies on the safety performance of the safety rating. In the study we collect safety rating and rainfall data at 15,732 reservoirs in Korea and then estimate statistical breach probabilities of the rainfall and safety rating groups. In 2020 the embankment breaches have occurred at 12 of B-rated, 10 of C-rated, and 3 of D-rated reservoirs. Under the extreme weathers it is found that the safety performance of the relatively good rated reservoirs is definitely outstanding than the safety performance of the relatively poor rated reservoirs.

Background/Objectives: A resonator plays a significant part in harnessing sound waves. An idea to use a resonator that only resonates on a specific frequency is hypothesized to harness sound efficiently. The primary objective of this study is to investigate the noise present in the boat engine room and be able to classify its resonant frequency and to know which of these frequencies occur most often. Methods/Statistical analysis: By using a sound recorder and a sound pressure level meter, the sound recorded from the idle state of the boat to disembarking and upon reaching the destination. Findings: Fast Fourier transform is applied to extract the frequencies of the recorded sound. Then the frequencies are evaluated using the Davies-Bouldin index to estimate the number of clusters and K-means clustering to classify the frequencies. Improvements/Applications: There are 2 clusters found and that Cluster 1 frequencies are from 17 Hz to 3295 Hz. Cluster 2 frequencies are from 6291 Hz to 8279 Hz. Among the available frequencies, 47 Hz, 155 Hz, and 186 Hz out weight all of the frequencies in terms of their occurrence, and among the three frequencies, 186 Hz has the highest amplitude.

Purpose: In this study, we investigated state-of-the-art lifelog analysis methods and describe a modeling process for analysis of acceleration data. These data can be easily obtained from smart devices and is useful for patients with visual display terminal (VDT) syndrome, who show specific lifestyle patterns. Methods and Results: We reviewed 23 recent lifelog articles focused on preprocessing, feature extraction, and classification to design a model for analysis of lifestyle patterns associated with VDT syndrome. Based on our review of articles, we recommend using relatively simple statistical indices, including min, max, median variance, and standard deviation, or frequency indices, such as power spectrum analysis for feature extraction. Based on favorable results with large datasets reported by several previous studies, we suggest using a genetic algorithm (GA) for classification. Notably, establishment of an organized human resource system for systematic data collection and management can improve data quality and also learning efficiency. Conclusion: We recommend the use of simple statistical indices and a GA for feature extraction and classification, respectively, to design a model for analysis of lifestyle patterns in patients with VDT syndrome. We also emphasize the importance of establishing an organized human resource system for systematic data collection and management to improve data quality and learning efficiency.

This study compares and analyzes the performance of LIME-based machine learning methods (Gaussian Naive Bayes (GNB), Highly-Efficient Logistic Regression (LR), Linear Support Vector Machine (SVM), and Triple-layer Neural Network (TNN)) using three medical datasets. High-dimensional data increases the likelihood of overfitting in learning algorithms due to the curse of dimensionality. To address this, LIME is utilized to compute the importance of key features contributing to the model's predictions. Based on this, features are selected. The LIME technique generates multiple samples by perturbing the data in the local region. Subsequently, a simple linear model is used to evaluate the impact of each feature on the predictions. Features with high importance derived from this process are selected for model retraining. As a result, it was confirmed that learning time could be reduced while maintaining or even improving performance with a smaller number of features. Consequently, by selecting necessary features, the curse of dimensionality issue is alleviated, and accuracy can be maintained or improved using fewer features in the Hepatitis C Prediction Dataset, Breast Cancer Wisconsin (Prognostic) Dataset, and Glioma Grading Clinical and Mutation Features Dataset.

목적: 본 연구는 실버세대 여성유명인을 대상으로 젊은 세대가 느끼는 이미지에 따라 유형화하고 유형에 따른 미용스타일과 이미지 차이를 분석하고자 한다. 이러한 연구는 새로운 소비계층으로 부각되고 있는 실버세대의 대표적인 스타일을 구분함으로써 실버세 대 여성에 대한 이해를 돕고 패션과 미용 상품 개발 및 이미지메이킹의 기초자료로 활용될 수 있을 것이다. 방법: 실버세대 여성유 명인을 55세 이상 65세 이하를 대상으로 50명을 선정하였으며, 이미지 평가를 위하여 선행연구를 참고로 총 20개의 형용사쌍을 추 출하였고 7점 의미미분척도로 구성하였다. 여성유명인의 선호도 평가는 '좋아하는-좋아하지 않는' 문항을 사용하였고 자료 수집은 2018년 6월 1일부터 6월 15일까지 이미지 평가를 실시하여 분석하였다. 결과: 첫째, 실버세대 여성유명인의 이미지 구성요인은 친 화성, 매력성, 활동성, 개성의 4개 요인으로 명명하였으며 전체 변량의 65.28%를 차지하였다. 둘째, 실버세대 여성유명인의 군집은 5개로 나타났으며 '부드러운 스타일', '차분한 스타일', '보수적인 스타일', '세련된 스타일', '사교적인 스타일'로 명명하였다. 셋째, 실 버세대 여성유명인의 유형에 따른 이미지 차이를 살펴본 결과, '부드러운 스타일'은 가장 친화성이 높았으며 '차분한 스타일'과 '세련 된 스타일'은 매력성이 두드러지고 '사교적인 스타일'은 활동성과 개성이 가장 높게 평가되었다. 넷째, 20대가 선호하는 실버세대 여 성유명인의 순위를 알아본 결과, 최하정, 김미숙 순으로 이들의 이미지는 매력적이고 개성이 두드러졌으며 선호도가 낮은 유명인은 김부선, 추미애 등으로 이들은 친화성이 낮게 나타났다. 결론: 실버세대 여성유명인의 이미지는 가장 중요한 차원으로 친화성 요인 임을 알 수 있으며, 군집에 따른 미용스타일의 분석된 특징을 참고하여, 추구하는 스타일 연출 및 이미지 구현에 활용해볼 수 있을 것이다. 또한, 친화적인 이미지로 평가 받고자 할 경우에는 '부드러운 스타일'로 연출하고, 매력적인 이미지를 나타내고자 한다면 '차 분한 스타일'과 '세련된 스타일'의 외형적 특징을 참고로 연출해 볼 수 있을 것이다. 활동적이며 개성이 뚜렷한 이미지를 고려한다면 '사교적인 스타일'로 연출하는 것이 효과적일 것이다.

Purpose: This study classified silver generation female celebrities according to images of the younger generation in order to analyze beauty, style, and image differences according to generation type. The study data can be used to inform fashion and beauty product development through improving our understanding of silver generation women's representative style, which is emerging as a new consumer segment. Methods: 50 images of silver age female celebrities who ranged from 55 to 65 years of age. A total of 20 adjective pairs were extracted with reference to a previous study of image evaluation using a 7-point semantic differential scale. The rating of female celebrities according to the adjectives was ‘like–dislike’. Data collection was conducted from June 1, 2018, to June 15, 2018. Results: The results of this study are summarized as follows. First, there were four factors constituting images of silver generation female celebrities—affinity, attractiveness, activity, and individuality—and accounted for 65.28% of the total variance. Second, the silver generation female celebrities were placed into one of five groups named ‘soft style’, ‘calm style’, ‘conservative style’, ‘refined style’, and ‘social style’. Third, as a result of examining the image difference according to type, ‘soft style’ had the highest affinity, ‘calm style’ and ‘refined style’ were rated as more attractive, and ‘sociable style’ had the highest activity and individuality rating. Fourth, the images of silver generation female celebrities deemed the most attractive and unique by those in their twenties were those of Hwa Jung Choi and Mi Sook Kim. Conclusion: The most important dimension of the silver generation female celebrity image was the affinity factor, which could be applied to the pursuit of style presentation and image implementation with reference to communitydefined characteristics of beauty and style. Accordingly, this study found that a friendly image was characterized by a ‘soft style’; an attractive image by the external features of a ‘calm style’ and/or ‘refined style’; and an active image with individuality was characterized by a ‘social style’.

目的: 根据年轻一代的形象对银发一族女性名人进行分类，以便根据世代类型分析美容，风格和形象差异。该 研究数据可用于通过提高我们对银代女性代表风格的理解来为时尚和美容产品的发展提供信息，该风格正在成 为新的消费群体。方法: 选定银发一族年龄从55岁到65岁之间的女性名人的50张照片。为评价形象参考先行研 究提取20个形容词对，并由7点语义差分异量表组成。女性名人的偏好评价为‘喜欢-不喜欢’。数据收集于2018 年6月1日至2018年6月15日进行。结果: 该研究的结果总结如下。首先，有四个因素构成银代女性名人的形象 - 亲和力，吸引力，活动和个性-占总变异的65.28％。其次，银代女性名人被分为五个群体之一，分别名为‘软风 格’，‘冷静风格’，‘保守风格’，‘精致风格’，‘社会风格’。第三，由于审查了 根据类型的图像差异，“柔和风格” 具有最高的亲和力，‘冷静风格’和‘精致风格’被评为更具吸引力，‘社交风格’具有最高的活动性和个性评级。 第 四，被二十几岁的人认为最具吸引力和独特的银代女性名人的形象是崔华贞和金美淑的形象。结论: 银发一族 女性名人形象最重要的维度是亲和力因素，可以应用于追求风格呈现和图像实现，参照社区定义的美和风格特 征。因此，这项研究发现友好形象的特点是‘柔和的风格’; 具有‘平静风格’和或‘精致风格’的外在特征的迷人形象; 具有个性的积极形象以‘社会风格’为特征。

In digital games, typography serves not only as a vehicle for conveying information but also as a crucial visual element that shapes the game’s identity and emotional atmosphere. However, prior research has predominantly focused on graphics, backgrounds, and character design, with systematic analyses of typographic expression remaining limited. This study concentrates on the emotional functions of typography in games by analyzing 25 PC games across five representative genres: role-playing (RPG), shooting (FPS/TPS), strategy (RTS/TBS), MOBA (AOS), and horror. The titles of these games were assessed using a seven-point scale based on typographic variables—weight, form, spacing, slant, baseline, and visual effects—and subsequently translated into emotional dimensions: robustness, stability, dynamism, traditionality, and fantasy. Based on this framework, genre-specific emotional typologies were identified. The results indicate that RPGs emphasize grandeur and mythic symbolism; FPS/TPS games highlight robustness and dynamism; strategy games exhibit order and stability; MOBAs convey competitive dynamism; and horror games strongly employ fantasy and anxiety. By classifying genre-specific emotional types of typography, this study expands the scope of game graphic design research to include textual expression. Practically, it provides design guidelines that help align typographic choices with genre-specific emotional characteristics. Nonetheless, the study is limited to PC games and a single-researcher evaluation, suggesting the need for future research to incorporate diverse platforms and user-based assessments.

Fuel structure is one of the major determinants in predicting the forest fire danger rating because it is a major factor in forest fire intensity and behavior. This study was conducted to the establishment of a fuel classification system by researching the fuel characteristics in a total of 12 study sites in Gangwon province, South Korea. There were two study sites in Yeongseo, and 10 study sites were in Yeongdong. We researched the fuel amount (kg/m2) separately in the study site by fuel structure in the forest. The fuel structure was divided into three types: surface layer (Duff, Litter, Twigs), shrub layer (the height of less than 2m), and crown layer (the height of 2m or more). As a result of this study, the fuel classification of 12 study sites was divided into three class by the K-means algorithm (Class Ⅰ: Chuncheon 1, Chuncheon 2, and Dogye-eup 1; Class Ⅱ: Jeo-dong 1, Jeo-dong 2, Yucheon-dong 1, Yucheon-dong 4, and Yucheon-dong 5; Class Ⅲ: Yucheon-dong 2, and Yucheon-dong 3).

본 연구에서는 사용자로부터 입력받은 이미지를 분석하여 그에 맞는 음악을 생성, 재생하는 방법을 고안 하였다. 단순히 이미지를 청각화 하는 기술적인 의미 뿐 아니라 사용자의 이미지에 담긴 정서와 의도 또한 담아내는 것을 목표로 하였다. 사용자는 본 연구에서 제안된 어플리케이션에 원하는 물체를 그린다. 인공지 능을 통해 이미지가 어떤 물체인지 판별 후, 그 물체와 이어질 수 있는 감정을 대응해 해당 멜로디의 감정과 분위기를 맞출 수 있도록 하였다. 정서에 알맞는 음정(key)를 설정한 뒤, 사용자가 이미지를 그릴 때 입력한 획순을 분석해 이를 기준으로 음계를 추출하여 선율을 생성하였다. 향후 이미지의 청각적 표현을 구현하는 것뿐만 아니라 그림에 대한 예술적인 이해와 의미 있는 음악을 만들어내기 위한 화성법 등의 작곡이론을 연 구하여 이미지에 담긴 예술성과 의도를 음악에 담아낼 수 있는 한 가지 방향을 제시할 것이다. 또한 그림을 인식하고 판별하기 위한 인공지능 기술과 그림 분석, 음악 생성 등의 예술 분야를 결합해 공학과 예술의 융 합이라는 방향으로서 의미 있는 시도가 될 것이다.

In this study, we devised a application that generates and reproduces music by analyzing images received from a user. It was aimed not only to capture the technical meaning of auditioning images, but also to express emotions and intentions in user's images. In the proposed application, a user draws a picture of a desired object. The application uses artificial intelligence to determine which object an image is. After that, the emotions that can be connected with each objects. The application determines the key that matches the mood through the emotion associated with the object. After setting a key suitable for the emotion, the user's stroke order is analyzed, and the melody is composed based on the extracted user’s stroke data. In the future, research on arts such as painting and music will be continued as well as implementing auditory expression of images. Based on this, we will present the direction to embody the artistic and intention in the image into music. It will also be a meaningful attempt as a direction of combination between engineering fields such as artificial intelligence for recognizing pictures and art fields such as picture analysis, and music production.

정부는 연구시설장비가 과학기술의 발전을 견인하는 매우 중요한 도구이자, 수단으로 여겨지면서 국 가적으로 R&D와 연구시설장비에 대한 예산 투자를 지속적으로 확대하였다. 또한, 기 구축된 국가연구 시설장비의 효율적 운영 및 체계적 관리의 필요성이 점차 대두되면서 2010년 12월, 국가연구시설장비 표준분류체계를 개발하였다. 그러나 연구현장에서는 국가연구시설장비의 NTIS(National Science and Technology Service) 정보수집 초기단계로 누적정보 부족에 따른 표준분류체계의 과학적 검증절차 부 재와 동일계층 간 분류기준의 비일관성 문제가 여전히 한계로 제기되고 있다. 따라서 본 연구는 지난 2010년, 2015년 각 제/개정된 국가연구시설장비 표준분류체계(대분류 8개, 중 분류 25개, 소분류 410개)의 분류 정확도를 측정하고자 선형판별분석(LDA)과 분산분석(ANOVA) 기법 을 적용하여 2단계로 분석하였다. 또한, 본 연구 분석을 위해 지난 10년 동안 NTIS에 누적 등록된 정 보데이터(Big-Data) 50,271건을 수집하여 이를 활용하였다. 이는 단순히 국내외 유사 분류체계와 전문 가 의견을 토대로 만들어진 現 국가연구시설 표준분류체계를 과학적으로 실증 검증한 첫 연구 사례에 해당된다. 본 연구 결과, 대분류 이하 중분류와 소분류로 분류된 개체 수의 집단별 판별정확도는 92.2% 로 매 우 높은 수준이었고, 분산분석을 통한 사후검증에서는 대분류 8개 중 2개 집단의 변별력이 다소 낮게 나타나, 現 표준분류체계 중 일부 개선이 필요한 것으로 조사되었다. 본 연구를 통해 現 국가연구시설 장비 표준분류체계가 향후 지속적으로 개선되길 바란다.

Recently, research F&E(Facilities and Equipment) have become very important as tools and means to lead the development of science and technology. The government has been continuously expanding investment budgets for R&D and research F&E, and the need for efficient operation and systematic management of research F&E built up nationwide has increased. In December 2010, The government developed and completed a standardized classification system for national research F&E. However, accuracy and trust of information classification are suspected because information is collected by a method in which a user(researcher) directly selects and registers a classification code in NTIS. Therefore, in the study, we analyzed linearly using linear discriminant analysis(LDA) and analysis of variance(ANOVA), to measure the classification accuracy for the standardized classification system(8 major-classes, 54 sub-classes, 410 small-classes) of the national research facilities and equipment established in 2010, and revised in 2015. For the analysis, we collected and used the information data(50,271 cases) cumulatively registered in NTIS(National Science and Technology Service) for the past 10 years. This is the first case of scientifically verifying the standardized classification system of the national research facilities and equipment, which is based on information of similar classification systems and a few expert reviews in the in-outside of the country. As a result of this study, the discriminant accuracy of major-classes organized hierarchically by sub-classes and small-classes was 92.2 %, which was very high. However, in post hoc verification through analysis of variance, the discrimination power of two classes out of eight major-classes was rather low. It is expected that the standardized classification system of the national research facilities and equipment will be improved through this study.

This bibliometric study is a citation analysis of journals that concern themselves with businessrelated topics. For this purpose, the journals from the following five categories from the Web of Science Subject Categories have been examined: “Business”, “Business & Finance”, “Economics”, “Management” and “Operations Research & Management Science”. The data is retrieved from the Journal Citation Reports 2019 for each journal that is part of one of the categories. The data includes information about the journals that the articles published in the specific journal in 2019 cited and which journals cited the articles in a specific journal (Cited Journal Data and Citing Journal Data). This data is combined by creating an asymmetrical 1- mode matrix of all journals. Then, the matrix is analyzed with Pajek and VOSviewer to create clusters of journals with a high inter-correlation through citations. In the following step, the journals in these clusters are analyzed for their disciplines, main topics, and compared to typical fields of study at universities (e.g. Accounting, Marketing, Finance, etc.) in order to highlight similarities and differences between the fields of studies and the clusters of journals.