• Journal of Intelligence and Information Systems
• /
• v.29 no.3
• /
• pp.419-437
• /
• 2023

As the number of thin filers in Korea surpasses 12 million, there is a growing interest in enhancing the accuracy of assessing their credit default risk to generate additional revenue. Specifically, researchers are actively pursuing the development of default prediction models using machine learning and deep learning algorithms, in contrast to traditional statistical default prediction methods, which struggle to capture nonlinearity. Among these efforts, Graph Neural Network (GNN) architecture is noteworthy for predicting default in situations with limited data on thin filers. This is due to their ability to incorporate network information between borrowers alongside conventional credit-related data. However, prior research employing graph neural networks has faced limitations in effectively handling diverse categorical variables present in credit information. In this study, we introduce the Transformer embedded Graph Convolutional Network (TeGCN), which aims to address these limitations and enable effective default prediction for thin filers. TeGCN combines the TabTransformer, capable of extracting contextual information from categorical variables, with the Graph Convolutional Network, which captures network information between borrowers. Our TeGCN model surpasses the baseline model's performance across both the general borrower dataset and the thin filer dataset. Specially, our model performs outstanding results in thin filer default prediction. This study achieves high default prediction accuracy by a model structure tailored to characteristics of credit information containing numerous categorical variables, especially in the context of thin filers with limited data. Our study can contribute to resolving the financial exclusion issues faced by thin filers and facilitate additional revenue within the financial industry.

• Journal of Service Research and Studies
• /
• v.13 no.1
• /
• pp.144-158
• /
• 2023

R&D (Research and Development) investment by companies and the government continues to increase, and the total R&D expenditure of public and private sectors in 2021 reached KRW 102 trillion, and the ratio of R&D investment to GDP was 4.96%. It ranks second in the world after Israel. This study analyzes the current status of the domestic R&D service industry and the research and development (R&D) innovation ecosystem based on the business activity survey data of the National Statistical Office in terms of improving R&D productivity of domestic R&D investment that has exceeded 100 trillion won, and analyzes the global R&D service industry and analyzes the global research service industry and R&D service Industry support government policy As a result of the analysis, in the domestic R&D ecosystem, the R&D outsourcing demand of service companies is rapidly increasing, compared to the stagnant R&D outsourcing demand of traditional manufacturing companies. In order to respond to these changing demands for technological innovation, this study suggests balanced support for R&D outsourcing for manufacturing and service companies, expansion of domestic R&D outsourcing, establishment of R&D outsourcing strategies specialized for the service industry, establishment of an open R&D ecosystem, and innovation in connection with materials/parts/equipment strategies.

• Journal of Korean Medicine Rehabilitation
• /
• v.33 no.4
• /
• pp.109-124
• /
• 2023

Objectives This study aims to find out effect of Korean medicine treatment on managing thoracolumbar compression fractures through retrospective observational study. Methods Among hospitalized patients at the Department of Korean Medicine Rehabilitation from January 1st, 2018 to February 28th, 2023, a total of 24 inpatients who were diagnosed with thoracolumbar compression fractures and received Korean medicine treatment were included in this study. Numeric rating scale (NRS) was used for pain assessment and clinical variables such as sex, symptoms, age, thoracolumbar injury classification and severity (TLICS) scores were collected. For subgroup analysis to analyze factors affecting treatment response, we divided patients into responders and non-responders according to NRS change. For statistical analysis, we compared before/after hospitalization and analyzed distinct features between two groups. Results Most of the patients were in their 70s and 83.33% were female. Average hospitalization period was 24.54±11.91 days. All patients had back pain as their chief complaint and only 2 patients received surgeries. In TLICS, only 1 patient got score 6, which represented surgery indication. After Korean medicine treatment, NRS of almost every patient got lower significantly at the time of discharge (3.02±1.93) than admission (5.52±1.95). Comparing two groups, responders had lower NRS at the time of discharge and TLICS score of them were lower than non-responders. Conclusions Our results show that Korean medicine treatment for thoracolumbar compression fractures was effective in reducing pain. There were distinct clinical features such as age, past history, surgeries between those with significant improvement in pain scores and those who did not.

• KIPS Transactions on Software and Data Engineering
• /
• v.13 no.1
• /
• pp.1-16
• /
• 2024

Over recent years, 6 Sigma has become a key methodology in manufacturing for quality improvement and cost reduction. However, challenges have arisen due to the difficulty in analyzing large-scale data generated by smart factories and its traditional, formal application. To address these limitations, a big data-based 6 Sigma approach has been developed, integrating the strengths of 6 Sigma and big data analysis, including statistical verification, mathematical optimization, interpretability, and machine learning. Despite its potential, the practical impact of this big data-based 6 Sigma on manufacturing processes and management performance has not been adequately verified, leading to its limited reliability and underutilization in practice. This study investigates the efficiency impact of DX SS, a big data-based 6 Sigma, on manufacturing processes, and identifies key success policies for its effective introduction and implementation in enterprises. The study highlights the importance of involving all executives and employees and researching key success policies, as demonstrated by cases where methodology implementation failed due to incorrect policies. This research aims to assist manufacturing companies in achieving successful outcomes by actively adopting and utilizing the methodologies presented.

• Journal of The Korean Association For Science Education
• /
• v.27 no.9
• /
• pp.787-795
• /
• 2007

The purpose of this study was to examine the effects of experimental learning using Small-Scale Chemistry (SSC) on science achievement and scientific attitude of high school students. SSC experiments were devised for 5 experiment themes of high school science textbook. Two classes were chosen from a high school in Busan and adopted into the comparison group and the experimental group; one group with thirty-four students participated in the class with an experimental learning using the SSC (experimental group), and another group with thirty-seven students participated in the class with the traditional learning (comparison group). The major discoveries of this study were as follows: Experimental learning using SSC has shown a significant difference between two groups in the science achievement of the students. Also there was a statistical difference between these two groups in the test which was conducted after a month to find out the durability of the experiment. Thus, the learning using SSC is assumed to be durable. Furthermore, there was a significant difference in scientific attitude between the two groups. It seemed that the learning using SSC had a good influence on fanning students' scientific attitude. In conclusion, an experimental learning using SSC has a positive effect on scientific achievement, the durability and scientific attitude of the students. We hope to develop suitable and various experimental learning materials using the SSC program that can be adopted in the classroom soon.

• Journal of Intelligence and Information Systems
• /
• v.22 no.3
• /
• pp.113-127
• /
• 2016

Among the techniques for recommendation, collaborative filtering (CF) is commonly recognized to be the most effective for implementing recommender systems. Until now, CF has been popularly studied and adopted in both academic and real-world applications. The basic idea of CF is to create recommendation results by finding correlations between users of a recommendation system. CF system compares users based on how similar they are, and recommend products to users by using other like-minded people's results of evaluation for each product. Thus, it is very important to compute evaluation similarities among users in CF because the recommendation quality depends on it. Typical CF uses user's explicit numeric ratings of items (i.e. quantitative information) when computing the similarities among users in CF. In other words, user's numeric ratings have been a sole source of user preference information in traditional CF. However, user ratings are unable to fully reflect user's actual preferences from time to time. According to several studies, users may more actively accommodate recommendation of reliable others when purchasing goods. Thus, trust relationship can be regarded as the informative source for identifying user's preference with accuracy. Under this background, we propose a new hybrid recommender system that fuses CF and social network analysis (SNA). The proposed system adopts the recommendation algorithm that additionally reflect the result analyzed by SNA. In detail, our proposed system is based on conventional memory-based CF, but it is designed to use both user's numeric ratings and trust relationship information between users when calculating user similarities. For this, our system creates and uses not only user-item rating matrix, but also user-to-user trust network. As the methods for calculating user similarity between users, we proposed two alternatives - one is algorithm calculating the degree of similarity between users by utilizing in-degree and out-degree centrality, which are the indices representing the central location in the social network. We named these approaches as 'Trust CF - All' and 'Trust CF - Conditional'. The other alternative is the algorithm reflecting a neighbor's score higher when a target user trusts the neighbor directly or indirectly. The direct or indirect trust relationship can be identified by searching trust network of users. In this study, we call this approach 'Trust CF - Search'. To validate the applicability of the proposed system, we used experimental data provided by LibRec that crawled from the entire FilmTrust website. It consists of ratings of movies and trust relationship network indicating who to trust between users. The experimental system was implemented using Microsoft Visual Basic for Applications (VBA) and UCINET 6. To examine the effectiveness of the proposed system, we compared the performance of our proposed method with one of conventional CF system. The performances of recommender system were evaluated by using average MAE (mean absolute error). The analysis results confirmed that in case of applying without conditions the in-degree centrality index of trusted network of users(i.e. Trust CF - All), the accuracy (MAE = 0.565134) was lower than conventional CF (MAE = 0.564966). And, in case of applying the in-degree centrality index only to the users with the out-degree centrality above a certain threshold value(i.e. Trust CF - Conditional), the proposed system improved the accuracy a little (MAE = 0.564909) compared to traditional CF. However, the algorithm searching based on the trusted network of users (i.e. Trust CF - Search) was found to show the best performance (MAE = 0.564846). And the result from paired samples t-test presented that Trust CF - Search outperformed conventional CF with 10% statistical significance level. Our study sheds a light on the application of user's trust relationship network information for facilitating electronic commerce by recommending proper items to users.

• Clinical and Experimental Pediatrics
• /
• v.46 no.2
• /
• pp.178-182
• /
• 2003

Purpose : Brain tumors are the second most common tumor in childhood, and medulloblastomas comprise 15-25% of brain tumors. The well known prognostic factors are age at diagnosis, stage of disease, and extent of surgical excision. In this study, we analysed the prognostic factors in patients who received chemotherapy after excision. Methods : We reviewed the medical records of 61 patients who received chemotherapy among the 94 patients who were diagnosed and treated between Jan 1985 and Sep 2001 in the Department of Pediatrics and Neurosurgery at Severance Hospital. Results : Among the total survival rate of patients who underwent chemotherapy, the 3-yr progression-free survival rate was $66.5{\pm}6.3%$ and the 15-yr progression-free survival rate was $60.3{\pm}6.7%$. The progression-free survival rate for patients with age at diagnosis over 3 yrs old and under 3 yrs old, was $64.5{\pm}7.7%$ and $48.2{\pm}12.9%$ respectively and there was no statistically significant difference. The survival rate of the high vs low risk group by staging was $72.7{\pm}10.5%$ and $54.6{\pm}8.3%$ respectively, and there was no significant difference. The survival rate of patients with total removal vs subtotal removal was $65.8{\pm}11.8%$ and $56.8{\pm}8.2%$ respectively, showing no statistical difference. Conclusion : The reason there is no difference in survival rate according to the traditional prognostic factors is that chemotherapy has improved not only the total survival rate but also the survival rate in patients with poor traditional prognostic factors. So, sufficient removal of tumor followed by proper chemotherapy and radiotherapy is an important factor which influences the survival rate of medulloblastoma patients.

• The Journal of Industrial Distribution & Business
• /
• v.1 no.1
• /
• pp.5-12
• /
• 2010

The retail business of drugstore was introduced to Korea for the first time 10 years ago. Since Olive Young introduced a retail store in the name of drugstore in 1999 for the first time in Korea, new distribution channel combining drugstore, cosmetic products and dairy products, etc has been made. At initial stage, the new distribution channel grew up slowly because of low specialty and economic stagnation. However, the three big distribution channels, that is to say, Olive Young (CJ), Watsons (GS) and W Store (Kolon Well Care), etc, were established to produce new distribution system following large-scaled discount stores as well as convenience stores. The purpose of the study is to investigate ways making Korean style drugstore be new retail business in addition to traditional markets, department stores, E-Mart and other general super markets and to examine problems preventing the drugstore from being promoted and to find out solutions. The speciality retailers that is called a category killer attacking department stores as well as marts is expanding market quickly. New consumption trend that gives priority to wellbeing is being expanded in accordance with high level of standards of living life: The drugstore is thought to be new alternative of distribution because it keeps special products. Young ladies who are main customers of drugstores respond to the trend sensitively to have more buying power that is thought to be promising. And, consumers' desire has become concrete and special. This is because consumers want not only convenient shopping but also special shopping system that is current trend. These days, so called Multi-shop and Total shop and other special shops have been recently opened. Special multi-shop has been concentrated on fashion product and miscellaneous goods so far: Health total wellbeing shop shall be popular in accordance with wellbeing trends. Drugstores can play an important role. Drugstores were opened for the first time ten years ago. In particular, Olive Young succeeded in going into the black after making efforts for a long time by many persons. Drugstores could succeed in the business owing to many persons in the past as well as customers who liked drugstores. However, drugstores once lost ways and recorded poor business results. The three drugstores, that is to say, Olive Young, Watsons making efforts to go into the black and W-Store pursuing traditional drugstore shall compete each other and make effort to satisfy customers' desire. In that way, the three drugstores can be assured of present business as well as future business. The consumers' demand trend has become special at sub-division so that drugstores that can satisfy the demand can succeed in the business. Large businesses may be more interested in the 4th generation retail business to produce good income and to have bright future. Drugstore business and market are likely to expand and develop owing to large business' participation in drugstore business. Drugstores expanded shop at Seoul and Gyeonggi-do until middle of 2000. Drugstore business at station sphere in Seoul and Gyeonggi-do that have high ratio of temporary population has low customer loyalty to have limitation on continuous growth. Since 2009, drugstores have opened new shops at local towns: From the year of 2010, drugstores need to establish multiple shop strategy by accelerating business speed and to allow customers to drop in the shop anywhere in the nation and to enter consumers' life deeply, so that they can strengthen business base definitely. Drugstores need to have price competitiveness to have multiple shop opening strategy and to satisfy consumers and to supply high quality services that is future subject to solve. And, Olive Young and Watsons that are Korean style drugstore need to keep system in order and to strengthen substance as Korean style drugstore and to expand marketing, so that they can get business outcome within 5 years that was done 10 years before and they become the 4th generation retail business. The study had difficulties at collecting material from the three drugstore because of poor cooperation. And, the author had great difficulty at collecting statistical material that was made in disorder. Further effort is needed considering such problems.

• Korean Journal of Agricultural Science
• /
• v.9 no.1
• /
• pp.260-268
• /
• 1982

The study has been carried out to investigate the relationship of certain factors to the private owners' attitudes toward forest management. The dependent variable is the attitude toward forest management, and the six independent variables are owner's age, educational level, forestland area, socioeconomic status, communicational and sociopsychological factors. The dependent variables were measured by Likert attitude scale. The data used in the study were obtained from 262 common-and model-forest owners, and analyzed primarily by multiple correlation statistical technique. Major findings of the study are summarized as follows; 1. The average score of common-owners' attitudes toward forest management is significantly lower than that of model-owners'. The scores of such independent variables as age, forestland area, socioeconomic status and communication of common-owners are also significantly lower than those of model-owners', while the significant difference of sociopsychological scores is not recognized between two owner groups. 2. The common-owners' attitudes toward forest management are negatively correlated with age variable, arid positively with other independent variables. On the other hand, the model-owners' are negatively correlated with education variable, arid positively with age, communicational and sociopsychological factors. 3. The multiple correlation coefficient between common-owner' management-attitude scores and six independent variables is 0.6971. About 41 percent of the variance of this attitude cart be explained by sociopsychological factor, and other 7 percent by communicational variable. In case of model-owners, the multiple correlation coefficient is 0.5207, and about 16.2 percent of the variance of the attitude can be explained by sociopsychological factor and 6 percent by age variable. 4. The multiple correlation coefficient between the attitude scores of all-owners arid such seven different personality factors as fatalistic, trustful, cooperative, ambitious, conservative, scientific and traditional familism personalities is 0.6566. In this case, 15 percent of the variance of the management-attitude can be explained by ambitious personality, and 11.2 percent by traditional familism. 5. The private forest owners' attitudes toward forest management are greatly affected by sociopsychological and communicational factors. The emphasis should be therefore put on the forestry extension works to induce private owners to management practices.

• The Korean Journal of Food And Nutrition
• /
• v.26 no.4
• /
• pp.737-744
• /
• 2013

The present study was investigated to evaluate on the health improving effect of Kochujang (Fermented Korean hot pepper paste), which was famous as Korean traditional healthy food among foreigners as it was served together with Bibimbap. For comparison, observation was made on the experimental animals (Sprague-Dawley rat) as breeding during 5 weeks, which were divided three groups as into the control groups (N-group) fed with normal feed only and high cholesterol feed only (H-group) and the K-group with combined diet of normal feed and Kochujang extracts (Water soluble complex materials). The most outstanding result was found as the definite fatty liver lesions found from all (100%) of the rats of the H-group at the abdominal opening examination. On the other hand, there was any finding on fatty liver lesions in the N-group as well as the K-group. Therefore, it was more than enough to identify the hypolipidemic effect by visual finding only. Furthermore, from the results of blood plasma analysis from SD-rats fed with Kochujang extract (K-group), more prominent differential findings were found, the value of HDL-cholesterol, which was good for preventing the arteriosclerosis, was more or less higher than that of H-groups. But the level of LDL-cholesterol, which was one of the hazardous agents of hyperlipidemia and arteriosclerosis, was more significantly reduced to the level even lower than that of N-group, based on the statistical analysis, in especial. According to the present results, we could find that the unidentified Korean traditional Kochujang-derived complex components had expressed both the positive health improving effects and hypolipidemic effects on the SD-rat.