• Science of Emotion and Sensibility
• /
• v.12 no.4
• /
• pp.433-442
• /
• 2009

People are tending to buy products through the Internet rather than purchasing them from the store. Some of the consumers give their feedback on line such as reviews, replies, comments, and blogs after they purchased the products. People are also likely to get some information through the Internet. Therefore, companies and public institutes have been facing this situation where they need to collect and analyze reviews or public opinions for them because many consumers are interested in other's opinions when they are about to make a purchase. However, most of the people's reviews on web site are too numerous, short and redundant. Under these circumstances, the emotion scanning system of text documents on the web is rising to the surface. Extracting writer's opinions or subjective ideas from text exists labeled words like GI(General Inquirer) and LKB(Lexical Knowledge base of near synonym difference) in English, however Korean language is not provided yet. In this paper, we labeled positive, negative, and neutral attribute at 4 POS(part of speech) which are noun, adjective, verb, and adverb in Korean dictionary. We extract construction patterns of emotional words and relationships among words in sentences from a large training set, and learned them. Based on this knowledge, comments and reviews regarding products are classified into two classes polarities with positive and negative using SO-PMI, which found the optimal condition from a combination of 4 POS. Lastly, in the design of the system, a flexible user interface is designed to add or edit the emotional words, the construction patterns related to emotions, and relationships among the words.

• KIPS Transactions on Software and Data Engineering
• /
• v.2 no.12
• /
• pp.873-880
• /
• 2013

This paper approaches the problem of resolving Korean zero pronouns using Centering Theory modeling local coherence. Centering Theory has been widely used to resolve English pronouns. However, it is much difficult to apply the centering framework for zero pronoun resolution in languages such as Japanese and Korean. Since in particular the use of non-anaphoric zero pronouns without explicit antecedents is not considered in the Centering Theory of Grosz et al., the presence of non-anaphoric cases negatively affects the performance of the resolution system based on Centering Theory. To overcome this, this paper presents a method which determines the intra-sentential anaphoricity of zero pronouns in subject position by using relationships between clauses, and then identifies antecedents of zero subjects. In our experiments, the proposed method outperforms the baseline method relying solely on Centering Theory.

• Informatization Policy
• /
• v.18 no.4
• /
• pp.103-137
• /
• 2011

The digital divide (DD) has been recognized as one of the most serious social and public policy issues to emerge from the rapid growth of computing facility, particularly in regard to Internet capabilities, over the past decade. In order to identify DD research trends and contributions, this research has adopted a bibliometric approach to quantitatively assess the research trends of the DD and also utilized information from related literature in the EV2 database from 2000 to 2011. The results show that the most frequently documented type of publications are journal articles, making up 55.2% of the total database of literature. English is the primary language (96.7%), and USA researchers have made the largest contribution (17.0%) to DD literature. Tokai University (Japan), Tilburg University (Netherlands), and Arizona State University (USA) rank as the top three author affiliations. Information technology, the Internet, and economic and social effects are the three top key words that appear in the related DD literature. The results also revealed various regional evidence, main topics, perspectives, frameworks, and solutions that were introduced in DD publications.

• The Journal of the Convergence on Culture Technology
• /
• v.8 no.2
• /
• pp.329-336
• /
• 2022

Unreadable Khitan scripts of ancient documents written by indian ink on parchment(sheepskin) are visualized by active infrared thermography without contacting and damaging the document which are deteriorated and aged presumably over many years. Sinusoidal infrared thermal wave using Halogen lamp is applied to the surface of the document in order to selectively magnify and record the thermal response of indian ink. The infrared image of the document captured in real time by infrared camera under the active external excitation shows the better sharpness and readability of Khitan characters than the optical image, from which many Khitan letters like ' ' and ' ' sounding as 'd' and 'ri' in English alphabet are detected and deciphered. It is concluded from the experiment that the active infrared thermography can be used as a promising method for digital reconstruction and preservation of ancient documents in the future.

• Journal of the Korea Convergence Society
• /
• v.13 no.3
• /
• pp.77-89
• /
• 2022

It has been proven through many previous researches that the pretrained language model with a large corpus helps improve performance in various natural language processing tasks. However, there is a limit to building a large-capacity corpus for training in a language environment where resources are scarce. Using the Cross-lingual Post-Training (XPT) method, we analyze the method's efficiency in Korean, which is a low resource language. XPT selectively reuses the English pretrained language model parameters, which is a high resource and uses an adaptation layer to learn the relationship between the two languages. This confirmed that only a small amount of the target language dataset in the relationship extraction shows better performance than the target pretrained language model. In addition, we analyze the characteristics of each model on the Korean language model and the Korean multilingual model disclosed by domestic and foreign researchers and companies.

• Journal of the Korea Furniture Society
• /
• v.21 no.4
• /
• pp.338-346
• /
• 2010

This study intends to gather terminology as a first step for encyclopedia of furniture from four areas such as traditional Korean furniture and architecture, modern and contemporary furniture, wood science, and wood culture by each expert. Since China, Japan, America, and European countries have chronologically influenced domestic furniture culture, the furniture terms are currently used by mixed languages such as Chinese characters, Japanese, English, Russian, Korean, etc., in the furniture fields and education. Domestic furniture companies and schools have had lots of difficulties in carrying out works due to the furniture terms not unified. This paper carried out through domestic furniture fields, internet, and furniture books from the inside and outside of the country, including the references of architecture, interior, crafts, and wood science and culture in order to research the terminology, results in the confusing language use of furniture terms at the furniture companies, schools for furniture education, and wood workshops having their own furniture terms. Therefore, for unified and standard furniture terminology, the terms gathered must be reviewed and conformed through the committee composed with the relevant experts so that it will be able to help the relevant business and education carry out their work better.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2018.05a
• /
• pp.693-695
• /
• 2018

In this paper, we propose the improvement of Hangul OCR accuracy through deep learning. OCR is a program that senses printed and handwritten characters in an optical way and encodes them digitally. In the case of the most commonly used Tesseract OCR, the accuracy of English recognition is high. However, Hangul has lower accuracy because it has less learning data for a complex structure. Therefore, in this study, we propose a method to improve the accuracy of Hangul OCR by extracting the character region from the desired image through image processing and using deep learning using it as learning data. It is expected that OCR, which has been developed only by existing alphanumeric and several languages, can be applied to various languages.

• Safety and Health at Work
• /
• v.5 no.2
• /
• pp.60-65
• /
• 2014

Background: To elucidate the relationship between safety culture maturity and safety performance of a particular company. Methods: To identify the factors that contribute to a safety culture, a survey questionnaire was created based mainly on the studies of $Fern{\acute{a}}ndez-Mu{\tilde{n}}iz$ et al. The survey was randomly distributed to 1000 employees of two oil companies and realized a rate of valid answer of 51%. Minitab 16 software was used and diverse tests, including the descriptive statistical analysis, factor analysis, reliability analysis, mean analysis, and correlation, were used for the analysis of data. Ten factors were extracted using the analysis of factor to represent safety culture and safety performance. Results: The results of this study showed that the managers' commitment, training, incentives, communication, and employee involvement are the priority domains on which it is necessary to stress the effort of improvement, where they had all the descriptive average values lower than 3.0 at the level of Company B. Furthermore, the results also showed that the safety culture influences the safety performance of the company. Therefore, Company A with a good safety culture (the descriptive average values more than 4.0), is more successful than Company B in terms of accident rates. Conclusion: The comparison between the two petrochemical plants of the group Sonatrach confirms these results in which Company A, the managers of which are English and Norwegian, distinguishes itself by the maturity of their safety culture has significantly higher evaluations than the company B, who is constituted of Algerian staff, in terms of safety management practices and safety performance.

• Journal of Engineering Education Research
• /
• v.11 no.1
• /
• pp.85-100
• /
• 2008

The purpose of this study is development of 'F1 in Schools Program' for elementary school student and validation of the program. Fifteen students were originally chosen from the W elementary school in Cheongju, Chungcheogbuk-Do. The children were tested on their understanding of science and 'F1 in Schools Program'. After developing the 'F1 in Schools Program', it was used in the classroom to evaluate applicability and to examine the responses from students, parents and schools. The program consists of 60 classes and the time of each class is 40 minutes. This study was conducted for 10 days from January 22nd, 2007 to February 2nd, 2007. CAD(Computer Aided Design), CFD(Computation Fluid Dynamics), CAM(Computer Aided manufacturing) and CNC Machine were used in this study and proved to be good materials for students in that they increased the students' participation and imagination. However, the children's cognitive and creative abilities as well as manuals written in English hampered the process. Most students, parents, schools seemed to be satisfied with use of the program. However, the schools showed that there was not enough understanding of the program as a whole. The processes with which students build and inspect using the $CO_2$Model Car not only improves the processing of the model but also enhanced the students scientific understanding related to the car speed.

• Journal of Information Management
• /
• v.38 no.3
• /
• pp.111-135
• /
• 2007

The purpose of this study is to identify the citing behaviors of researchers in the field of mechanical engineering. It tries to verify if there is a significant difference on citing behavior of researchers between the past and the present with dissertations produced in P University as samples. For the comparison, years 1996 and 2004 are selected for the citation analysis. It analyzed four aspects, such as types of resources cited, languages used in cited documents, years since their publication of cited documents, and the journals indexed in SCI. The results of the analysis are; First, journals are the most cited than any other types of information resources. The citation of WWW resources which are gradually increasing for research is not shown in 1996, but there were some cited in 2004. Second, doctoral candidates usually cite document in English for their study. Statistics show that the use of resources in Japanese is on the decrease. Third, doctoral candidates in the discipline prefer materials published within 4-7 years, 8-11 years rather than 0-3 years since their publication. Last, journals indexed in SCI among the citation in dissertations are about 33 percent for both 1996 and 2004.