• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.21 no.1
• /
• pp.169-177
• /
• 2020

In South Korea, the results of R&D in science and technology are submitted to the National Science and Technology Information Service (NTIS) in reports that have Korea national science and technology standard classification codes (K-NSCC). However, considering there are more than 2000 sub-categories, it is non-trivial to choose correct classification codes without a clear understanding of the K-NSCC. In addition, there are few cases of automatic document classification research based on the K-NSCC, and there are no training data in the public domain. To the best of our knowledge, this study is the first attempt to build a highly performing K-NSCC classification system based on NTIS report meta-information from the last five years (2013-2017). To this end, about 210 mid-level categories were selected, and we conducted preprocessing considering the characteristics of research report metadata. More specifically, we propose a convolutional neural network (CNN) technique using only task names and keywords, which are the most influential fields. The proposed model is compared with several machine learning methods (e.g., the linear support vector classifier, CNN, gated recurrent unit, etc.) that show good performance in text classification, and that have a performance advantage of 1% to 7% based on a top-three F1 score.

• Cartoon and Animation Studies
• /
• s.44
• /
• pp.85-116
• /
• 2016

This is the in-depth research dealing with Animatoon, the only animation-specialized magazine in Korea. By examining articles provided under the 'Animatoon Report' category, which is one of article categories set by the publisher, this research tries to find out the topic of articles and examine the tendency of topic. 'Animatoon Report' is chosen because the category title does not clearly show characteristics of articles, but has the second largest number of articles among all categories. First, the articles of the first ten years (1995-2005) has the largest number of pages as well as images, which means each article of this period tries to enrich its contents. Second, the role of magazine reporters are not critical considering the number of articles written by reporters are smaller than those made by the editorial department. Third, articles tries to deal with diverse issues and are mostly placed in front of the magazine. Fourth, in early days, articles used to provide the 3~5 lines of introductory summaries, but changed into provide the lead, the Korean and English subtitle, and so on. Fifth, articles mostly focus on issues of animation and Korea rather than other areas and countries. The results on the basis of article type are as follows. Among the types of people, work, organization/company, event, policy/industry, and etc, the policy/industry type has the largest amount of articles, and articles on the policy and those on the industry shows similar number of articles. Second, the event type has the second largest number of articles even though there are several separate categories only for the event. Third, articles in the et cetera type are often seen in early days because Animatoon who did not systemize the company itself focused on animation history and production techniques. Fourth, articles both on the people and the work type are consistently seen through the whole time period, but those on organization/company and event has more and more articles as time passes. In conclusion, it is possible to mention that the 'Animatoon Report' type shows more interest in policy and industry, and its interest on both issues are consistently seen from 2001 to 2015.

• Journal of Intelligence and Information Systems
• /
• v.20 no.3
• /
• pp.77-92
• /
• 2014

Recently, numerous documents including unstructured data and text have been created due to the rapid increase in the usage of social media and the Internet. Each document is usually provided with a specific category for the convenience of the users. In the past, the categorization was performed manually. However, in the case of manual categorization, not only can the accuracy of the categorization be not guaranteed but the categorization also requires a large amount of time and huge costs. Many studies have been conducted towards the automatic creation of categories to solve the limitations of manual categorization. Unfortunately, most of these methods cannot be applied to categorizing complex documents with multiple topics because the methods work by assuming that one document can be categorized into one category only. In order to overcome this limitation, some studies have attempted to categorize each document into multiple categories. However, they are also limited in that their learning process involves training using a multi-categorized document set. These methods therefore cannot be applied to multi-categorization of most documents unless multi-categorized training sets are provided. To overcome the limitation of the requirement of a multi-categorized training set by traditional multi-categorization algorithms, we propose a new methodology that can extend a category of a single-categorized document to multiple categorizes by analyzing relationships among categories, topics, and documents. First, we attempt to find the relationship between documents and topics by using the result of topic analysis for single-categorized documents. Second, we construct a correspondence table between topics and categories by investigating the relationship between them. Finally, we calculate the matching scores for each document to multiple categories. The results imply that a document can be classified into a certain category if and only if the matching score is higher than the predefined threshold. For example, we can classify a certain document into three categories that have larger matching scores than the predefined threshold. The main contribution of our study is that our methodology can improve the applicability of traditional multi-category classifiers by generating multi-categorized documents from single-categorized documents. Additionally, we propose a module for verifying the accuracy of the proposed methodology. For performance evaluation, we performed intensive experiments with news articles. News articles are clearly categorized based on the theme, whereas the use of vulgar language and slang is smaller than other usual text document. We collected news articles from July 2012 to June 2013. The articles exhibit large variations in terms of the number of types of categories. This is because readers have different levels of interest in each category. Additionally, the result is also attributed to the differences in the frequency of the events in each category. In order to minimize the distortion of the result from the number of articles in different categories, we extracted 3,000 articles equally from each of the eight categories. Therefore, the total number of articles used in our experiments was 24,000. The eight categories were "IT Science," "Economy," "Society," "Life and Culture," "World," "Sports," "Entertainment," and "Politics." By using the news articles that we collected, we calculated the document/category correspondence scores by utilizing topic/category and document/topics correspondence scores. The document/category correspondence score can be said to indicate the degree of correspondence of each document to a certain category. As a result, we could present two additional categories for each of the 23,089 documents. Precision, recall, and F-score were revealed to be 0.605, 0.629, and 0.617 respectively when only the top 1 predicted category was evaluated, whereas they were revealed to be 0.838, 0.290, and 0.431 when the top 1 - 3 predicted categories were considered. It was very interesting to find a large variation between the scores of the eight categories on precision, recall, and F-score.

• Journal of Hospice and Palliative Care
• /
• v.14 no.2
• /
• pp.91-100
• /
• 2011

Purpose: This study was performed to investigate patients' preferences on receiving life-sustaining treatments (LST) and to analyze the relationship between patients' characteristics and LST selection. We also examined any discrepancy between LST patients' choices regarding medical intervention and actual medical intervention given/not given within 48 hours before death. Methods: This cross-sectional study was performed from March 1, 2008 to August 31, 2008 in the Palliative Care Unit of Korea University Hospital. Electric medical records (EMR) of 102 hospice cancer patients were reviewed, and 74 patients with Glasgow coma scale (GCS) ${\geq}$10 at the time of signing the advance medical directives (AMD) were selected for the first analysis. Then, patients alive at the end of this study, transferred to other hospitals or dead within 48 hours were excluded, and the remaining 42 patients were selected for the second analysis. Results: Preferred LST included antibiotics, total parenteral nutrition, tube feeding, transfusion, and laboratory and imaging studies. The relationship between patients' characteristics and LST could not be analyzed due to skewed preferences. LST chosen at the time of signing the AMD and actual medical intervention given/not given in the last 48 hours showed discrepancy in most cases. Conclusion: When making AMD in hospice cancer patients, it is important to consider the time and possibility of changing the choices. Above all, patients must fully understand the AMD. Thus, LST should always be provided with careful consideration of all possibilities, because legal and social aspects of AMD have not been established yet.

• Journal of radiological science and technology
• /
• v.39 no.4
• /
• pp.651-660
• /
• 2016

Due to recent population aging, the number of check-up for senior citizens has increased steadily. According to this trend, the market size of dental X-ray equipment and the number of approval and review for these devices have simultaneously increased. The technical document of medical device is required for approval and review for medical device, and medical device companies needs to have work comprehension and expertise, as the document needs to include the overall contents such as performances, test criteria, etc.. Yet, since most of domestic manufacturers or importers of medical devices are small businesses, it is difficult for them to recruit professional manpower for approval of medical devices, and submission of inaccurate technical documents has increased. These problems lead to delay of the approval process and to difficulties in quick entering into the market. Especially, the Ministry of Food and Drug safety (MFDS) standards of a dental extra-oral X-ray equipment, a dental intra-oral X-ray equipment, an arm-type computed tomography, and a portable X-ray system have been recently enacted or not. this guideline of dental X-ray equipment adjusting revised standards was developed to help relative companies and reviewers. For this study, first, the methods to write technical document have been reviewed with revised international and domestic regulations and system. Second, the domestic and foreign market status of each item has been surveyed and analyzed. Third, the contents of technical documents already approved by MFDS have been analyzed to select the correct example, test items, criteria, and methods. Finally, the guideline has been developed based on international and domestic regulation, through close review of a consultative body composed of academic, industrial, research institute and government experts.

• Economic and Environmental Geology
• /
• v.48 no.6
• /
• pp.509-523
• /
• 2015

This study introduces newly implemented geological well logs database for Jeju public water wells, built for a research project focusing on integrated hydrogeology database of Jeju Island. A detailed analysis of the existing 1,200 Jeju Island geological logs for the public wells developed since 1970 revealed six major indications to be improved for their use in Jeju geological logs DB construction: (1) lack of uniformity in rock name classification, (2) poor definitions of pyroclastic deposits and sand and gravel layers, (3) lack of well borehole aquifer information, (4) lack of information on well screen installation in many water wells, (5) differences by person in geological logging descriptions. A new Jeju geological logs DB enabling standardized input and output formats has been implemented to overcome the above indications by reestablishing the names of Jeju volcanic and sedimentary rocks and utilizing a commercial, database-based input structured, geological log program. The newly designed database structure in geological log program enables users to store a large number of geology, well drilling, and test data at the standardized DB input structure. Also, well borehole groundwater and aquifer test data can be easily added without modifying the existing database structure. Thus, the newly implemented geological logs DB could be a standardized DB for a large number of Jeju existing public wells and new wells to be developed in the future at Jeju Island. Also, the new geological logs DB will be a basis for ongoing project 'Developing GIS-based integrated interpretation system for Jeju Island hydrogeology'.

• Korean Journal of Heritage: History & Science
• /
• v.52 no.4
• /
• pp.142-153
• /
• 2019

The Joseon Dynasty (1392-1910) left behind a wealth of documentary heritage, including collections of literary works, personal letters, and journals, as well as public documents such as Veritable Records of the Joseon Dynasty (Joseon Wangjo Sillok), Diaries of the Royal Secretariat (Seungjeongwon Ilgi), and State Protocols (Uigwe). Such heritage also includes personal diaries that have been highly regarded for their frank and vivid records of people's lives in the Joseon period. There have been great diaries published and intended for reading by the Korean public, including War Diaries (Nanjung Ilgi, 1592~1598) by Yi Sun-sin and Diaries of Jehol (Yeolha Ilgi, 1780) by Park Ji-won. Unfortunately, a great majority of these personal records remain unknown to the world. Such great records have not been given an opportunity to be documented properly, but are left outside public attention, abandoned to be damaged and destroyed. Few personal diaries of the Joseon period were written on good-quality paper. After the death of their authors, these diaries were left to be kept by their descendants; this explains why many of these records have been in poor condition, particularly when compared with the public records published by the government of Joseon, such as Sillok and Uigwe, even when these were lucky enough to be taken care of by the authors' descendants. Even after surviving a long time, many of these personal records remain in the form of manuscripts, written in semi-cursive and cursive scripts of Chinese characters, thus making it even more difficult for the people of the current generation -- most of whom have not been given an opportunity to learn Chinese characters at school -- to take care of their documentary heritage properly. Meanwhile, it is also true that, as the value of the public records published by the government of Joseon as historical materials has grown, they are used more often as content for TV dramas such as Daejanggeum. At the same time, there have been increasingly louder voices citing the need for the study, preservation, and management of the personal diaries from Joseon. Considering the situation, this study provides a general overview of the personal diaries of Joseon as recently surveyed by the National Research Institute of Cultural Heritage, as well as their characteristic features, subjects, and backgrounds. This study is expected to contribute to future research on the preservation and management of the personal diaries of Joseon.