• (The)Study of the Eastern Classic
• /
• no.37
• /
• pp.33-61
• /
• 2009

In the late Koryo period, until 14th century, there had been at least two groups of Manchurians who were conferred citizenships; one group was living as an original inhabitant in the coastal area of north­eastern part of Korean peninsular, long time ago, and they were over one thousand households. The other was coming down from inland, eastern part of Yoha River, to the area of Tuman River to settle down and they were at least around one hundred and sixty households, including such tribes as Al-tha-ry, Ol-lyang-hap, Ol-jok-hap and others. They were treated courteously, from the early days of Choson dynasty, with governmental policies in an economic, political, and social ways. They were given, for instance, a house, a land, household furniture, and clothes. They were allowed to get marry with a native Korean to settle down. They were educated how to cultivate their lands. It was also possible for them to be given an official position politically or allowed to take a National Civil Official Examination. The fact they could take such an Examination, in particular, means they were treated fairly and equally, because they also had a privilege to improve their social positions through the formal system as much as common people. Two typical families were scrutinized, in this paper, family Chong-hae Lee and family Chon-ju Ju. All of them were successful to settle down with different backgrounds each other. The former were from a headman, Lee Jee-ran, who controlled his tribe, over five hundred households. He was given three titles of a meritorious retainer at the founding of Chosun dynasty, at the retrieval of armies, and an enshrined retainer. His son, Lee Wha-yong, was also given a vassal of merit who kept a close tie successfully with the king's family through a marriage. Upon the foundation of their ancestors, their grandsons, family Lee Hyo-yang and family Lee Hyo-gang, each, had taken solid root as an aristocratic Yang-ban class. The former became a high officer family, generation by generation, while the latter changed into a civil official family through Civil Official Examinations. They lived mainly around Seoul, Kyong-gi Province and some lived in their original places, Ham-kyong Province. Chu-man, the first ancestor, was given a meritorious retainer at the founding of the dynasty and Chu-in was also given a high officer position from the government. They kept living at the original place, Ham-heung, Ham-kyong Province, and then became an outstanding local family there. They began to pass the Civil Official Examinations. After 17th century on the passers were 17 in Civil Official Examinations and 40 were passed in lower civil examinations. The positions in government they attained usually were remonstrance which position was prohibited particularly to North­Western people at that time. The Chosun dynasty was open to Machurians widely through the system of envoy, convoy, and naturalization. It was intended to build up an enclosure policy through a friendly diplomatic relation with them against any possible invasion from outside. This is one reason why they were supported fully that much in a various way.

• Korean Journal of Psychosomatic Medicine
• /
• v.24 no.1
• /
• pp.102-108
• /
• 2016

Objectives : The deficit of recognition memory has been found as one of the common neurocognitive impairments in patients with schizophrenia. In addition, they were reported to fail to enhance the memory about emotional stimuli. Previous studies have shown that bilateral eye movements enhance the memory retrieval. Therefore, this study was conducted in order to investigate the memory enhancement of bilaterally alternating eye movements in schizophrenic patients. Methods : Twenty one patients with schizophrenia participated in this study. The participants learned faces (angry or neutral faces), and then performed a recognition memory task in relation to the faces after bilateral eye movements and central fixation. Recognition accuracy, response bias, and mean response time to hits were compared and analysed. Two-way repeated measure analysis of variance was performed for statistical analysis. Results : There was a significant effect of bilateral eye movements condition in mean response time(F=5.812, p<0.05) and response bias(F=10.366, p<0.01). Statistically significant interaction effects were not observed between eye movement condition and face emotion type. Conclusions : Irrespective of the emotional difference of facial stimuli, recognition memory processing was more enhanced after bilateral eye movements in patients with schizophrenia. Further study will be needed to investigate the underlying neural mechanism of bilateral eye movements-induced memory enhancement in patients with schizophrenia.

• Korean Journal of Remote Sensing
• /
• v.28 no.1
• /
• pp.11-19
• /
• 2012

Air-borne pollen, biogenically created aerosol particle, influences Earth's radiative balance, visibility impairment, and human health. The importance of pollens has resulted in numerous experimental studies aimed at characterizing their dispersion and transport, as well as health effects. There is, however, limited scientific information concerning the optical properties of airborne pollen particles contributing to total ambient aerosols. In this study, for the first time, optical characteristics of pollen such as aerosol backscattering coefficient, aerosol extinction coefficient, and depolarization ratio at 532 nm and their effect to the atmospheric aerosol were studied by lidar remotes sensing technique. Dual-Lidar observations were carried out at the Gwangju Institute of Science & Technology (GIST) located in Gwagnju, Korea ($35.15^{\circ}E$, $126.53^{\circ}N$) for a spring pollen event from 5 to 7 May 2009. The pollen concentration was measured at the rooftop of Gwangju Bohoon hospital where the building is located 1.0 km apart from lidar site by using Burkard trap sampler. During intensive observation period, high pollen concentration was detected as 1360, 2696, and $1952m^{-3}$ in 5, 6, and 7 May, and increased lidar return signal below 1.5km altitude. Pollen optical depth retrieved from depolarization ratio was 0.036, 0.021, and 0.019 in 5, 6, and 7 May, respectively. Pollen particles mainly detected in daytime resulting increased aerosol optical depth and decrease of Angstrom exponent.

• Journal of KIISE:Computing Practices and Letters
• /
• v.14 no.4
• /
• pp.412-429
• /
• 2008

As the amount of electronic documents increases rapidly with the growth of the Internet, a parallel search engine capable of handling a large number of documents are becoming ever important. To implement a parallel search engine, we need to partition the inverted index and search through the partitioned index in parallel. There are two methods of partitioning the inverted index: 1) document-identifier based partitioning and 2) keyword-identifier based partitioning. However, each method alone has the following drawbacks. The former is convenient in inserting documents and has high throughput, but has poor performance for top h query processing. The latter has good performance for top-k query processing, but is inconvenient in inserting documents and has low throughput. In this paper, we propose a hybrid partitioning method to compensate for the drawback of each method. We design and implement a parallel search engine that supports the hybrid partitioning method using the Odysseus DBMS tightly coupled with information retrieval capability. We first introduce the architecture of the parallel search engine-Odysseus/parallel-OOSQL. We then show the effectiveness of the proposed system through systematic experiments. The experimental results show that the query processing time of the document-identifier based partitioning method is approximately inversely proportional to the number of blocks in the partition of the inverted index. The results also show that the keyword-identifier based partitioning method has good performance in top-k query processing. The proposed parallel search engine can be optimized for performance by customizing the methods of partitioning the inverted index according to the application environment. The Odysseus/parallel OOSQL parallel search engine is capable of indexing, storing, and querying 100 million web documents per node or tens of billions of web documents for the entire system.

• Clinical and Experimental Reproductive Medicine
• /
• v.26 no.2
• /
• pp.171-178
• /
• 1999

The follicular fluid (FF) of ovary contains various biological active products which affected on the growth of follicles and the fertilization of oocyte in physiological reproductive process of mammals. This study was designed to determine the effects of human FF on fertilization of oocyte and embryonal development in vitro culture. The FF was prepared as clear without blood contamination by needle aspiration from mature follicles of human at the time of oocytes retrieval for in vitro fertilization (IVF). As the medium for culture in vitro of embryonal cells, human tubal fluid (HTF) supplemented with follicular fluids at concentrations of 10%, 40% and pure FF were used. These effects were compared to control group of cultured embryos in HTF supplemented with 0.4% BSA (bovine serum albumin). For IVF, 64 eggs in control group, 67 eggs in 10% FF, 57 eggs in 40% FF and 64 eggs in pure FF were respectively allocated. And the rates of fertilization were almost similar in all groups as resulting 82.81% in control, 85.07% in 10% FF, 87.71% in 40% FF and 81.25% in pure FF. On the examination for embryonal cleavage from fertilized eggs, the rates of developing to 4 cell stage was similar in all groups, as results 98.11% in control, 98.27% in 10% FF and 98% in 40% FF but 78.84% in pure FF. And the rates of developing to 8-16 cell stage were significantly reduced as 44% in 40% FF and 44.23% in pure FF (p<0.05) compare to 71.69% in control media. As likewise, the rates of developing to morular stage were also significantly reduced to 36% (p<0.05) and 21.15% (p<0.01) respectively in 40% FF and pure FF. And the rates to blastocystic stage of embryo was lowest as 7.69% in pure FF (Table 1). The quality of embryonal cells on cleavage to the 8-16 cell stage was poorer, higher concentrations of FF. The rates of grade 1 in pure FF, as 23.07%, was lowest compare to those of other groups, in which the rates of grade 1 in control, 10% FF and 40% FF were 58.49%, 47.36% and 34% respectively. And on the contrary, the rate of grade 4 in pure FF was highest as 23.07%, while those were 5.66% in control, 8.77% in 10% FF and 20% in 40% FF (Table 2). On the viability of embryos, the rate of embryonal cell death was more rise, at the higher concentrations as well as longer exposure in the follicular fluid. At 48 hours after in vitro culture of embryos, the rate of survival embryos in pure FF was markedly lowered as 44.23%, compare to that of control (p<0.05). But there was not significant difference between the rates of survival embryos in each group beside the pure FF, which the rates were 77.35% in control, 70.17% in 10% FF and 60% in 40% FF respectively. And at 72 hours after in vitro culture, the rates of survival embryos were also significantly dropped to 21.15% in pure and 36% in 40% at concentration of FF compare to 62.26% in control (p<0.05, p<0.01). Finally, the rate of embryonal death at 96 hours after in vitro culture was highest as 82.69% in pure FF among all groups which those were 35.84 in control, 56.14% in 10% FF and 64% in 40% FF respectively (Fig. 1, 2, 3). In conclusion, this study suggests that the FF has no effects, in particular, to the in vitro fertilization of oocytes but exerted a bad effect to the cleavage, quality and viability of the embryonal cells during in vitro culture. However, the FF is harmful on embryonal development at conditions in higher concentration and especially on the embryos after $8{\sim}16$ cell stage.

• Journal of Intelligence and Information Systems
• /
• v.25 no.1
• /
• pp.43-61
• /
• 2019

Development of technologies in artificial intelligence has been rapidly increasing with the Fourth Industrial Revolution, and researches related to AI have been actively conducted in a variety of fields such as autonomous vehicles, natural language processing, and robotics. These researches have been focused on solving cognitive problems such as learning and problem solving related to human intelligence from the 1950s. The field of artificial intelligence has achieved more technological advance than ever, due to recent interest in technology and research on various algorithms. The knowledge-based system is a sub-domain of artificial intelligence, and it aims to enable artificial intelligence agents to make decisions by using machine-readable and processible knowledge constructed from complex and informal human knowledge and rules in various fields. A knowledge base is used to optimize information collection, organization, and retrieval, and recently it is used with statistical artificial intelligence such as machine learning. Recently, the purpose of the knowledge base is to express, publish, and share knowledge on the web by describing and connecting web resources such as pages and data. These knowledge bases are used for intelligent processing in various fields of artificial intelligence such as question answering system of the smart speaker. However, building a useful knowledge base is a time-consuming task and still requires a lot of effort of the experts. In recent years, many kinds of research and technologies of knowledge based artificial intelligence use DBpedia that is one of the biggest knowledge base aiming to extract structured content from the various information of Wikipedia. DBpedia contains various information extracted from Wikipedia such as a title, categories, and links, but the most useful knowledge is from infobox of Wikipedia that presents a summary of some unifying aspect created by users. These knowledge are created by the mapping rule between infobox structures and DBpedia ontology schema defined in DBpedia Extraction Framework. In this way, DBpedia can expect high reliability in terms of accuracy of knowledge by using the method of generating knowledge from semi-structured infobox data created by users. However, since only about 50% of all wiki pages contain infobox in Korean Wikipedia, DBpedia has limitations in term of knowledge scalability. This paper proposes a method to extract knowledge from text documents according to the ontology schema using machine learning. In order to demonstrate the appropriateness of this method, we explain a knowledge extraction model according to the DBpedia ontology schema by learning Wikipedia infoboxes. Our knowledge extraction model consists of three steps, document classification as ontology classes, proper sentence classification to extract triples, and value selection and transformation into RDF triple structure. The structure of Wikipedia infobox are defined as infobox templates that provide standardized information across related articles, and DBpedia ontology schema can be mapped these infobox templates. Based on these mapping relations, we classify the input document according to infobox categories which means ontology classes. After determining the classification of the input document, we classify the appropriate sentence according to attributes belonging to the classification. Finally, we extract knowledge from sentences that are classified as appropriate, and we convert knowledge into a form of triples. In order to train models, we generated training data set from Wikipedia dump using a method to add BIO tags to sentences, so we trained about 200 classes and about 2,500 relations for extracting knowledge. Furthermore, we evaluated comparative experiments of CRF and Bi-LSTM-CRF for the knowledge extraction process. Through this proposed process, it is possible to utilize structured knowledge by extracting knowledge according to the ontology schema from text documents. In addition, this methodology can significantly reduce the effort of the experts to construct instances according to the ontology schema.

• Journal of Intelligence and Information Systems
• /
• v.27 no.3
• /
• pp.157-173
• /
• 2021

With the development of information technology, the amount of available information increases daily. However, having access to so much information makes it difficult for users to easily find the information they seek. Users want a visualized system that reduces information retrieval and learning time, saving them from personally reading and judging all available information. As a result, recommendation systems are an increasingly important technologies that are essential to the business. Collaborative filtering is used in various fields with excellent performance because recommendations are made based on similar user interests and preferences. However, limitations do exist. Sparsity occurs when user-item preference information is insufficient, and is the main limitation of collaborative filtering. The evaluation value of the user item matrix may be distorted by the data depending on the popularity of the product, or there may be new users who have not yet evaluated the value. The lack of historical data to identify consumer preferences is referred to as data sparsity, and various methods have been studied to address these problems. However, most attempts to solve the sparsity problem are not optimal because they can only be applied when additional data such as users' personal information, social networks, or characteristics of items are included. Another problem is that real-world score data are mostly biased to high scores, resulting in severe imbalances. One cause of this imbalance distribution is the purchasing bias, in which only users with high product ratings purchase products, so those with low ratings are less likely to purchase products and thus do not leave negative product reviews. Due to these characteristics, unlike most users' actual preferences, reviews by users who purchase products are more likely to be positive. Therefore, the actual rating data is over-learned in many classes with high incidence due to its biased characteristics, distorting the market. Applying collaborative filtering to these imbalanced data leads to poor recommendation performance due to excessive learning of biased classes. Traditional oversampling techniques to address this problem are likely to cause overfitting because they repeat the same data, which acts as noise in learning, reducing recommendation performance. In addition, pre-processing methods for most existing data imbalance problems are designed and used for binary classes. Binary class imbalance techniques are difficult to apply to multi-class problems because they cannot model multi-class problems, such as objects at cross-class boundaries or objects overlapping multiple classes. To solve this problem, research has been conducted to convert and apply multi-class problems to binary class problems. However, simplification of multi-class problems can cause potential classification errors when combined with the results of classifiers learned from other sub-problems, resulting in loss of important information about relationships beyond the selected items. Therefore, it is necessary to develop more effective methods to address multi-class imbalance problems. We propose a collaborative filtering model using CGAN to generate realistic virtual data to populate the empty user-item matrix. Conditional vector y identify distributions for minority classes and generate data reflecting their characteristics. Collaborative filtering then maximizes the performance of the recommendation system via hyperparameter tuning. This process should improve the accuracy of the model by addressing the sparsity problem of collaborative filtering implementations while mitigating data imbalances arising from real data. Our model has superior recommendation performance over existing oversampling techniques and existing real-world data with data sparsity. SMOTE, Borderline SMOTE, SVM-SMOTE, ADASYN, and GAN were used as comparative models and we demonstrate the highest prediction accuracy on the RMSE and MAE evaluation scales. Through this study, oversampling based on deep learning will be able to further refine the performance of recommendation systems using actual data and be used to build business recommendation systems.

• Korean Journal of Heritage: History & Science
• /
• v.54 no.2
• /
• pp.136-153
• /
• 2021

The restitution of cultural heritage located abroad has been going on for 107 years, starting with the return of the Jigwangguksa Tower to Beopcheonsaji Temple in Wonju after it was taken during the Japanese occupation in 1915. The Overseas Cultural Heritage Foundation, established in 2012, has laid the foundation for retrieval, preservation, restoration, and exchange of cultural heritage through research cooperation and the purchasing of cultural heritage items. The pace of the collection of cultural heritage objects and the locating of others has increased every year since its establishment, and the number of returned, rather than recovered, cultural heritage items has also increased. The present study aimed to complete a value inquiry of the cultural relics of Waryongmae (臥龍梅) and a restitution of cultural heritage as the main focuses. The process of recovering relics from Waryongmae has been recorded in the book The Cultural Property Returned into Our Arms, published by the above-mentioned foundation. This record was revised and supplemented to try and raise its cultural value by adding elaborate storytelling to the process of recovering the Waryongmae that grew in the courtyard of Changdeokgung Palace. The cultural value of Waryongmae is that it is unique. The Waryongmae is the first living cultural heritage, and therefore has cultural value due to its uniqueness. Second, the Waryongmae has unique cultural value due to its restitution and return to Korea twice, once in 1992, and another time in 1999. The first restitution was special in that it was featured by the Japanese media, and the second was special in that it was intensively reported by the Korean media. Third, 42 Waryongmae cultural content types were explored, including nineteen visual contents, eleven interactive contents, and twelve skate contents.

• Journal of the Korean earth science society
• /
• v.41 no.6
• /
• pp.617-629
• /
• 2020

Sea surface temperature (SST), which plays an important role in climate change and global environmental change, can be divided into skin sea surface temperature (SSST) observed by satellite infrared sensors and the bulk temperature of sea water (BSST) measured by instruments. As sea surface temperature products distributed by many overseas institutions represent temperatures at different depths, it is essential to understand the relationship between the SSST and the BSST. In this study, we constructed an observation system of infrared radiometer onboard a marine research vessel for the first time in Korea to measure the SSST. The calibration coefficients were prepared by performing the calibration procedure of the radiometer device in the laboratory prior to the shipborne observation. A series of processes were applied to calculate the temperature of the layer of radiance emitted from the sea surface as well as that from the sky. The differences in skin-bulk temperatures were investigated quantitatively and the characteristics of the vertical structure of temperatures in the upper ocean were understood through comparison with Himawari-8 geostationary satellite SSTs. Comparison of the skin-bulk temperature differences illustrated overall differences of about 0.76℃ at Jangmok port in the southern coast and the offshore region of the eastern coast of the Korean Peninsula from 21 April to May 6, 2020. In addition, the root-mean-square error of the skin-bulk temperature differences showed daily variation from 0.6℃ to 0.9℃, with the largest difference of 0.83-0.89℃ at 1-3 KST during the daytime and the smallest difference of 0.59℃ at 15 KST. The bias also revealed clear diurnal variation at a range of 0.47-0.75℃. The difference between the observed skin sea surface temperature and the satellite sea surface temperature showed a mean square error of approximately 0.74℃ and a bias of 0.37℃. The analysis of this study confirmed the difference in the skin-bulk temperatures according to the observation depth. This suggests that further ocean shipborne infrared radiometer observations should be carried out continuously in the offshore regions to understand diurnal variation as well as seasonal variations of the skin-bulk SSTs and their relations to potential causes.

• Journal of Intelligence and Information Systems
• /
• v.24 no.3
• /
• pp.1-19
• /
• 2018

Large amount of data is now available for research and business sectors to extract knowledge from it. This data can be in the form of unstructured data such as audio, text, and image data and can be analyzed by deep learning methodology. Deep learning is now widely used for various estimation, classification, and prediction problems. Especially, fashion business adopts deep learning techniques for apparel recognition, apparel search and retrieval engine, and automatic product recommendation. The core model of these applications is the image classification using Convolutional Neural Networks (CNN). CNN is made up of neurons which learn parameters such as weights while inputs come through and reach outputs. CNN has layer structure which is best suited for image classification as it is comprised of convolutional layer for generating feature maps, pooling layer for reducing the dimensionality of feature maps, and fully-connected layer for classifying the extracted features. However, most of the classification models have been trained using online product image, which is taken under controlled situation such as apparel image itself or professional model wearing apparel. This image may not be an effective way to train the classification model considering the situation when one might want to classify street fashion image or walking image, which is taken in uncontrolled situation and involves people's movement and unexpected pose. Therefore, we propose to train the model with runway apparel image dataset which captures mobility. This will allow the classification model to be trained with far more variable data and enhance the adaptation with diverse query image. To achieve both convergence and generalization of the model, we apply Transfer Learning on our training network. As Transfer Learning in CNN is composed of pre-training and fine-tuning stages, we divide the training step into two. First, we pre-train our architecture with large-scale dataset, ImageNet dataset, which consists of 1.2 million images with 1000 categories including animals, plants, activities, materials, instrumentations, scenes, and foods. We use GoogLeNet for our main architecture as it has achieved great accuracy with efficiency in ImageNet Large Scale Visual Recognition Challenge (ILSVRC). Second, we fine-tune the network with our own runway image dataset. For the runway image dataset, we could not find any previously and publicly made dataset, so we collect the dataset from Google Image Search attaining 2426 images of 32 major fashion brands including Anna Molinari, Balenciaga, Balmain, Brioni, Burberry, Celine, Chanel, Chloe, Christian Dior, Cividini, Dolce and Gabbana, Emilio Pucci, Ermenegildo, Fendi, Giuliana Teso, Gucci, Issey Miyake, Kenzo, Leonard, Louis Vuitton, Marc Jacobs, Marni, Max Mara, Missoni, Moschino, Ralph Lauren, Roberto Cavalli, Sonia Rykiel, Stella McCartney, Valentino, Versace, and Yve Saint Laurent. We perform 10-folded experiments to consider the random generation of training data, and our proposed model has achieved accuracy of 67.2% on final test. Our research suggests several advantages over previous related studies as to our best knowledge, there haven't been any previous studies which trained the network for apparel image classification based on runway image dataset. We suggest the idea of training model with image capturing all the possible postures, which is denoted as mobility, by using our own runway apparel image dataset. Moreover, by applying Transfer Learning and using checkpoint and parameters provided by Tensorflow Slim, we could save time spent on training the classification model as taking 6 minutes per experiment to train the classifier. This model can be used in many business applications where the query image can be runway image, product image, or street fashion image. To be specific, runway query image can be used for mobile application service during fashion week to facilitate brand search, street style query image can be classified during fashion editorial task to classify and label the brand or style, and website query image can be processed by e-commerce multi-complex service providing item information or recommending similar item.