• Journal of Intelligence and Information Systems
• /
• v.25 no.4
• /
• pp.105-122
• /
• 2019

Dimensionality reduction is one of the methods to handle big data in text mining. For dimensionality reduction, we should consider the density of data, which has a significant influence on the performance of sentence classification. It requires lots of computations for data of higher dimensions. Eventually, it can cause lots of computational cost and overfitting in the model. Thus, the dimension reduction process is necessary to improve the performance of the model. Diverse methods have been proposed from only lessening the noise of data like misspelling or informal text to including semantic and syntactic information. On top of it, the expression and selection of the text features have impacts on the performance of the classifier for sentence classification, which is one of the fields of Natural Language Processing. The common goal of dimension reduction is to find latent space that is representative of raw data from observation space. Existing methods utilize various algorithms for dimensionality reduction, such as feature extraction and feature selection. In addition to these algorithms, word embeddings, learning low-dimensional vector space representations of words, that can capture semantic and syntactic information from data are also utilized. For improving performance, recent studies have suggested methods that the word dictionary is modified according to the positive and negative score of pre-defined words. The basic idea of this study is that similar words have similar vector representations. Once the feature selection algorithm selects the words that are not important, we thought the words that are similar to the selected words also have no impacts on sentence classification. This study proposes two ways to achieve more accurate classification that conduct selective word elimination under specific regulations and construct word embedding based on Word2Vec embedding. To select words having low importance from the text, we use information gain algorithm to measure the importance and cosine similarity to search for similar words. First, we eliminate words that have comparatively low information gain values from the raw text and form word embedding. Second, we select words additionally that are similar to the words that have a low level of information gain values and make word embedding. In the end, these filtered text and word embedding apply to the deep learning models; Convolutional Neural Network and Attention-Based Bidirectional LSTM. This study uses customer reviews on Kindle in Amazon.com, IMDB, and Yelp as datasets, and classify each data using the deep learning models. The reviews got more than five helpful votes, and the ratio of helpful votes was over 70% classified as helpful reviews. Also, Yelp only shows the number of helpful votes. We extracted 100,000 reviews which got more than five helpful votes using a random sampling method among 750,000 reviews. The minimal preprocessing was executed to each dataset, such as removing numbers and special characters from text data. To evaluate the proposed methods, we compared the performances of Word2Vec and GloVe word embeddings, which used all the words. We showed that one of the proposed methods is better than the embeddings with all the words. By removing unimportant words, we can get better performance. However, if we removed too many words, it showed that the performance was lowered. For future research, it is required to consider diverse ways of preprocessing and the in-depth analysis for the co-occurrence of words to measure similarity values among words. Also, we only applied the proposed method with Word2Vec. Other embedding methods such as GloVe, fastText, ELMo can be applied with the proposed methods, and it is possible to identify the possible combinations between word embedding methods and elimination methods.

• Restorative Dentistry and Endodontics
• /
• v.34 no.3
• /
• pp.169-176
• /
• 2009

Diabetes Mellitus (DM) is a syndrome accompanied with the abnormal secretion or function of insulin, a hormone that plays a vital role in controlling the blood glucose level (BGL). Type land 2 DM are most common form and the prevalence of the latter is recently increasing, The aim of this article was to assess whet her Type 2 DM could act as a predisposing risk factor on the pulpo-periapical pathogenesis. Previous literature on the pathologic changes of blood vessels in DM was thoroughly reviewed. Furthermore, a histopathologic analysis of artificially-induced periapical specimens obtained from Type 2 diabetic and DM-resistant rats was compared. Histopathologic results demonstrate that the size of periapical bone destruction w as larger and the degree of pulpal inflammation was more severe in diabetic rats, indicating that Type 2 D M itself can be a predisposing risk factor that makes the host more susceptible to pulpal infection. The possible reasons may be that in diabetic state the lumen of pulpal blood vessels are thickened by atheromatous deposits, and microcirculation is hindered, The function of polymorphonuclear leukocyte is also impair ed and the migration of immune cells is blocked, leading to increased chance of pulpal infection. Also, lack of collateral circulation of pulpal blood vessels makes the pulp more susceptible to infection. These decrease the regeneration capacity of pulpal cells or tissues, delaying the healing process, Therefore, when restorative treatment is needed in Type 2 DM patients, dentists should minimize irritation to the pulpal tissue un der control of BGL.

• Korean Journal of Heritage: History & Science
• /
• v.53 no.2
• /
• pp.4-23
• /
• 2020

The purpose of this study was to explore the diversity of Korea's iron casting technology and to examine various casting methods. The study involved a literature review, analysis of artifacts, local investigation of production tools and technology, and scientific analysis of casting and cast materials. Bellows technology, or Bulmi technology, is a form of iron casting technology that uses bellows to melt cast iron before the molten iron is poured into a clay cast. This technology, handed down only in Jeju Island, relies on use of a clay cast instead of the sand cast that is more common in mainland Korea. Casting methods for cast iron pots can be broadly divided into two: sand mold casting and porcelain casting. The former uses a sand cast made from mixing seokbire (clay mixed with soft stones), sand and clay, while the latter uses a clay cast, formed by mixing clay with rice straw and reed. The five steps in the sand mold casting method for iron pot are cast making, filling, melting iron into molten iron, pouring the molten iron into the cast mold, and refining the final product. The six steps in the porcelain clay casting method are cast making, cast firing, spreading jilmeok, melting iron into molten iron, pouring the molten iron, and refining the final product. The two casting methods differ in terms of materials, cast firing, and spreading of jilmeok. This study provided insight into Korea's unique iron casting technology by examining the scientific principles behind the materials and tools used in each stage of iron pot casting: collecting and kneading mud, producing a cast, biscuit firing, hwajeokmosal (building sand on the heated cast) and spreading jilmeok, drying and biyaljil (spreading jilmeok evenly on the cast), hapjang (combining two half-sized casts to make one complete cast), producing a smelting furnace, roasting twice, smelting, pouring molten iron into a cast, and refining the final product. Scientific analysis of the final product and materials involved in porcelain clay casting showed that the main components were mud and sand (SiO₂, Al₂O₃, and Fe₂O₃). The release agent was found to be graphite, containing SiO₂, Al₂O₃, Fe₂O₃, and K₂O. The completed cast iron pot had the structure of white cast iron, comprised of cementite (Fe₃C) and pearlite (a layered structure of ferrite and cementite).

• Childhood Kidney Diseases
• /
• v.17 no.2
• /
• pp.101-109
• /
• 2013

Purpose: Pediatric urolithiasis is uncommon in children but is a cause of significant morbidity and damage to the kidney. Although much information on adult urolithiasis is available in the literature, large studies on the pediatric population are still scarce. In this report, we review our experience with pediatric urolithiasis over 22 years at a tertiary referral center. Method: We retrospectively reviewed the records of children with newly diagnosed urolithiasis between January 1991 and May 2013. We assessed the age, sex, family history, initial symptoms, location of stones, underlying cause, stone analysis, treatment, and recurrence among the patients. Results: In total, 137 patients (96 male, 41 female) were assessed. The age range was 0-17 years (mean age, 6.0 years). Forty-three (31%) children were aged <1 year, and 37% (16/43) had a history of intensive care unit (ICU) admission. Thirteen patients (9.5%) had a family history of stones. The most common symptoms at presentation among the patients were gross hematuria (56/137, 41%) and flank or abdominal pain (46/137, 34%). The stones were located in the kidney (85/137, 62%), ureter (29/137, 21%), bladder (2/137, 1.4%), and multiple locations (20/137, 15 %). Congenital abnormalities of the genitourinary (G-U) tract, with or without metabolic abnormality, or urinary tract infection (UTI) was detected in 26 children (19%). Ninety-one patients (66%) underwent metabolic examination, and 38% of these patients exhibited an abnormality. UTI, with or without abnormalities of the G-U tract, or metabolic abnormality was detected in 26 children (19%). Of the 35 stones analyzed, the majority were calcium stones (20/35, 57%), followed by infected stones (5/35, 14%), uric acid stones (4/35, 11%), carbonate apatite stones (3/35, 7%), cystine stones (2/35, 6%), and phosphate stones (1/35, 3%). Five patients (4%) required open procedures, with or without non-open procedures, whereas 77 patients (56%) were managed conservatively; the remaining 55 patients (40%) received some other form of intervention. Eighteen patients (13%) had stone recurrence during the follow-up period. Conclusions: Pediatric urolithiasis is commonly associated with abnormalities of the G-U tract and/or metabolic disorders and/or UTI. Half of the patients will pass their stones spontaneously, and all the techniques of minimally invasive surgery are applicable in the treatment of children with stones. As the recurrence rates are high among this population, long-term follow-up is recommended and the complete clearance of stones is important.

• The Korean Journal of Nuclear Medicine Technology
• /
• v.21 no.1
• /
• pp.65-69
• /
• 2017

Purpose KSNMT(Korea Society of Nuclear Medicine Technology) stepping first step in 1997, has published first journal related with nuclear medicine technology in 1985. With classifying In Vivo Session Dissertation reported in the entire journal, trend of the Dissertation will be studied. Materials and Methods Dissertations which published from 1985 to first half of 2016 in the journal are classified with presentation form and with scanner, And all the data is organized with Excel program. Through the data, the number of dissertations published in each year, the number of dissertation published in details, and keyword distributions in each period are analyzed. Results The number of In-vivo section dissertations was 1151 and the number of In-vivo section dissertations that have common subject with In-vitro section was 28. The number of In-vivo section dissertation in 1980s was 46, in 1990s was 149, in 2000 was 467 and from 2010 to the first half of 2016 was 517. The number of dissertation with original articles was 571, with abstract was 529, with symposium was 31, with special lecture was 25, with review was 11, with interesting image was 7, with poster was 3 and with case report was 2. With symposium and special lecture excluded, which count 56, the number of dissertation with PET was 319, with Planar was 302, with SPECT was 172, with radiopharmaceutical was 113, with guard and safety management 103, with BMD was 28, etc. was 86. The number of dissertation about oncology was 201, about scanner was 179, about cardiovascular and circulatory system was 102, about safe environment was 82, about musculoskeletal system was 76, about nervous nuclear medicine was 66, about quality assurance was 61, about genitourinary system was 56, about endocrine system was 49, about digestive system was 44, about Therapy, about industrial safety was 24, about molecular imaging was 15, infection and inflammation was 9, about respiratory system was 8 and etc. was 108. The mostly used keyword through 1999 to 2005 was PET and through 2006 to 2016 was PET/CT. Conclusion To encourage various dissertations to be submitted, Korea Society of Nuclear Medicine should analyze date about not only about dissertations that are already published, but also about various research materials. Moreover, Korea Society of Nuclear Medicine also have to provide technical support such as sharing big data from homepage and systematical support to its member to publish dissertation that has high impact factor. It is important each individual researcher to have continuing effort as well as each organization cooperation.

• Journal of Intelligence and Information Systems
• /
• v.22 no.2
• /
• pp.127-142
• /
• 2016

Deep learning model is a kind of neural networks that allows multiple hidden layers. There are various deep learning architectures such as convolutional neural networks, deep belief networks and recurrent neural networks. Those have been applied to fields like computer vision, automatic speech recognition, natural language processing, audio recognition and bioinformatics where they have been shown to produce state-of-the-art results on various tasks. Among those architectures, convolutional neural networks and recurrent neural networks are classified as the supervised learning model. And in recent years, those supervised learning models have gained more popularity than unsupervised learning models such as deep belief networks, because supervised learning models have shown fashionable applications in such fields mentioned above. Deep learning models can be trained with backpropagation algorithm. Backpropagation is an abbreviation for "backward propagation of errors" and a common method of training artificial neural networks used in conjunction with an optimization method such as gradient descent. The method calculates the gradient of an error function with respect to all the weights in the network. The gradient is fed to the optimization method which in turn uses it to update the weights, in an attempt to minimize the error function. Convolutional neural networks use a special architecture which is particularly well-adapted to classify images. Using this architecture makes convolutional networks fast to train. This, in turn, helps us train deep, muti-layer networks, which are very good at classifying images. These days, deep convolutional networks are used in most neural networks for image recognition. Convolutional neural networks use three basic ideas: local receptive fields, shared weights, and pooling. By local receptive fields, we mean that each neuron in the first(or any) hidden layer will be connected to a small region of the input(or previous layer's) neurons. Shared weights mean that we're going to use the same weights and bias for each of the local receptive field. This means that all the neurons in the hidden layer detect exactly the same feature, just at different locations in the input image. In addition to the convolutional layers just described, convolutional neural networks also contain pooling layers. Pooling layers are usually used immediately after convolutional layers. What the pooling layers do is to simplify the information in the output from the convolutional layer. Recent convolutional network architectures have 10 to 20 hidden layers and billions of connections between units. Training deep learning networks has taken weeks several years ago, but thanks to progress in GPU and algorithm enhancement, training time has reduced to several hours. Neural networks with time-varying behavior are known as recurrent neural networks or RNNs. A recurrent neural network is a class of artificial neural network where connections between units form a directed cycle. This creates an internal state of the network which allows it to exhibit dynamic temporal behavior. Unlike feedforward neural networks, RNNs can use their internal memory to process arbitrary sequences of inputs. Early RNN models turned out to be very difficult to train, harder even than deep feedforward networks. The reason is the unstable gradient problem such as vanishing gradient and exploding gradient. The gradient can get smaller and smaller as it is propagated back through layers. This makes learning in early layers extremely slow. The problem actually gets worse in RNNs, since gradients aren't just propagated backward through layers, they're propagated backward through time. If the network runs for a long time, that can make the gradient extremely unstable and hard to learn from. It has been possible to incorporate an idea known as long short-term memory units (LSTMs) into RNNs. LSTMs make it much easier to get good results when training RNNs, and many recent papers make use of LSTMs or related ideas.

• Journal of Christian Education in Korea
• /
• v.67
• /
• pp.45-75
• /
• 2021

The purpose of this study is to explore a new paradigm for Bible didactics in the context of the contemporary times and it turns its gaze to the midrash, the old tradition of Hebrew Bible interpretation. In order for the current Bible study to be meaningful and effective in today's situation, it is an effort to connect the Bible and us well, more than educational contents or materials. The word "midrash" itself means "textual interpretation", or "study", derived from the root verb darash, which means "to seek," "to seek with care," "to enquire," "to require" forms of which appear frequently in the Hebrew Bible. Midrash means an exegesis and interpretation of the Hebrew Bible (Torah) as well as a group of works that are the result of specific interpretations of the rabbis. This rabbinical tradition provides specific interpretative guidelines dealing with the Bible. These interpretive guidelines were passed down and formed an attitude of interpreting the Bible that is still relevant today. The rabbinical interpretative guidelines in midrash lead to the discovery of the following biblical didactical meanings. First, the Bible requires an attitude of listening and learning. Second, an attitude of inquiry is needed. Third, an exploration through the empty space is essential. Fourth, it leads us to recognize the importance of mutual respect and communication. Fifth, through the Words that challenge me, the meaning of biblical teaching is discovered. These interpretation guidelines have much in common with Bibliodrama, which applies midrash to the didactic of Bible. Bibliodrama is a dramatic inquiry, where the effect of in-depth inquiry and consideration that midrash aimed at can be expected. In addition, bibliodrama is a process of communal interaction that leads to a new experience and a richer understanding of the Bible through different positions and viewpoints. Exploring the "white fire" of the Bible, we listen to what God says to us, which causes us to change and form an identity. The biblical didactical meaning found in midrash's interpretation guidelines and the biblical didactical application of midrash through the bibliodrama can be presented as a new alternative to Christian education for the past, the present and the future. This will be able to present a new paradigm for biblical didactics with the word of God living and working in the present, not the Bible of the past, which is far from our present life.

• 기호학연구
• /
• no.57
• /
• pp.7-26
• /
• 2018

The purpose of this paper is to interpret the discourse on design thinking through the perspective of Roland Barthes' Mythologies. To this end, this paper will explore the mythologization process of design thinking using the methodological framework of Barthes, which structurally interprets the connotations produced using semiosis. Design thinking originally refers to a method which is used in the process of planning ideas about designs in order to create the final products for professional designs. However, design thinking has recently attracted more interest from the public because it has become known as a tool for solving various problems which exist outside of the field of design, such as social issues, management, and marketing strategies. Barthes points out that myths are used as a tool to deliver ideologies. He also emphasizes the importance of 'structural thinking'. It interprets the inherent connotative meanings more than the denotative meanings, which are explicitly shown. One of the most powerful ideologies which our society embraces today is creativity. Design thinking realizes the manifestation of creativity through a schematized process. This can be explained by considering design thinking as an icon that is specifically turned into a figuration to realize its objectness, in which a discourse for solving issues and social codes meet together and form a mythology. The mythologies that Barthes cites in his book refer to mythical values created by the cultural codes which humans have produced in our modern and contemporary age. The symbolic value of design thinking has become more important than the signifier which design thinking itself presents. This means that design thinking has become a sign that has mythical properties. In other words, the ideology of creativity embodied by design thinking has attained a mythological status, as it produces a new cultural code through innovation. The process of interpreting a phenomenon using the perspective of semiotics is an important tool that allows us to examine the concept of an object and its surroundings thoroughly. This paper attempts to expand the external scope of critical analysis about social phenomena by using the signs which continuously reveal themselves in common ideologies, such as design thinking, which has been gaining more popularity recently.

• (The) Research of the performance art and culture
• /
• no.37
• /
• pp.61-94
• /
• 2018

This article examined that the historical changes and current significance of the Jaedam(재담) and the Xiangsheng(相聲), one of the traditional Korean and Chinese language games. Both Korean Jaedam and Chinese Xiangsheng are representative language games and traditional performing arts for laughing. The origin of the Jaedam can be traced back to Uheui(우희). Uheui has been called Changyouxi in China, Bae Woo-hee, and Jo Hee in Korea. Uheui is the most traditional language game and a variety of performances were derived from its spreading and inheriting process. Among them, Korean Jaedam and Chinese Xiangsheng can be said to be a piece of art that has successfully inherited Uheui tradition. From the late 18th century, Korean Jaedam were established as independent performance arts, and became highly active in many performance by professional joker Park Chun-jae and other performers. With the development of gramophone record in the early 20th century, the Jaedam was mainly made on the theater stage and radio. At this time, the new performance art of 'Mandam(만담)' was derived from the Jaedam, which focused more on satire current events and criticizing the social situation. Mandam has been popular for a long time and then extinct in the 21st century. The jaedam have been handed down only in the Korean traditional performance so far. Meanwhile, Chinese Xiangsheng, which was built in the mid-19th century, a bit later than Korean Jaedam, was initially considered to be a vulgar art of the lower class, but finally became popular in the early 20th century. In the mid-20th century, Xiangsheng was transformed into a new character, which mainly deals with social praise and edification of the masses. But since 'New Xiangsheng' does not focous on a satire on social conditions, the humor has been reduced. In the early 21st century, Xiangsheng was on the verge of extinction just like Mandam, but through the efforts of young actors to revive tradition, another reformation of this art was made to return to tradition and small theater. Currently, the 'traditional Xiangsheng', which has returned to tradition, is once again receiving the love and support of the Chinese audience. Korean Jaedam and Chinese Xiangsheng have many similarities in terms of history and recruitment, but they are now in different fates. There is also a great deal in common ground in terms of the content and form of the two arts. In the case of Xiangsheng, it is one of the traditional folk art forms which is still loved by the Chinese people and has become one of the most important traditional performances. On the other hand, in Korea, Jaedam as independent performance arts has disappeared and now only can be seen in traditional performances such as 'Korean mask theater'. The fact that Korean Jaedam and Chinese Xiangsheng have undergone similar changes in their spreading and inheriting process, while Korean Jaedam have disappeared and Chinese Xiangsheng is well preserved. The reason can be confirmed through the main idea of this article.

• Korean Journal of Heritage: History & Science
• /
• v.47 no.1
• /
• pp.116-133
• /
• 2014

The memorial stones of Sumanotap were known to have been sequentially produced in 1719, 1773, 1874, 1653, 1874. However, after examining the chronological era (年號) and sexagesimal cycle (干支), it was found that the second and fourth memorial stones were both recorded in 1713. The reason why the records of the same period were divided into two is assumed that since the subject to repair was divided into main structure and flagpole of the temple, and it was recorded per each monk in charge and donator. On the fifth tablet of the memorial stone of Sumanotap contains all the names of monks who were in charge of repairing Sumanotap as well as the names of common donators. These people are not verified in detail though, through comparing the Buddhist names on the third memorial stone with Buddhist paintings and painting records, 12 monks were identified to be in the same names. Generally, they were active in Gangwon province, Gyeongsang province, Seoul and Gyeonggi province in the late 19th century and even after the Sumanotap repair, they took up other duties of various temples. The most notable name among the donators recorded in the memorial stones of Sumanotap was Kim Jwageun. He was the donator for the chest of gold and silver that was enshrined in the pagoda in 1874. His name was recorded with Mrs. Yang from Cheongsin and, at that time, was known just as a donator. However, actually, the donator was not Kim Jwageun but his mistress Mrs. Yang who used to be a gisaeng, called Na-hap and was from Naju. Kim Jwageun deceased in 1869 already. It is exceptional that Mrs. Yang who were the mistress of the prime minister Kim Jwageun personally participated in the Buddhist event in which king, queen, mothers of king and queen, noblewoman participated. Besides, the event of Sumanotap repair was carried out by royal family in order to celebrate the hundredth day of the prince, Sunjong and his grandfather, Heungseon Daewongun did not participate for political reason. In this situation, the domination of Mrs. Yang is considered remarkable. Meanwhile, the temples that frequently appeared among the 10 temples recorded in the memorial stone of Sumanotap are Gakhwa-sa temple in Bonghwa-gun, Gyeongsangbuk-do and Bodeok-sa temple in Yeongwolgun, Gangwon-do. Gakhwa-sa temple was a temple where the historical achievements were kept so was under the government's control. Bodeok-sa temple was the buddhist temple praying for the soul of the Danjong buried in the tomb and was managed by the government. Therefore, the face these two temples participated in the Sumanotap repair proves that Sumanotap was repaired by the supports form these two temples and the stature of Sumanotap at that time was as high as the temples where the historical achievements are store or the royal buddhist temples for praying for ancestors' souls.