• Journal of the Korea Society of Computer and Information
• /
• v.26 no.1
• /
• pp.57-67
• /
• 2021

Recently, multi-modal deep learning techniques that combine heterogeneous data for deep learning analysis have been utilized a lot. In particular, studies on the synthesis of Text to Image that automatically generate images from text are being actively conducted. Deep learning for image synthesis requires a vast amount of data consisting of pairs of images and text describing the image. Therefore, various data augmentation techniques have been devised to generate a large amount of data from small data. A number of text augmentation techniques based on synonym replacement have been proposed so far. However, these techniques have a common limitation in that there is a possibility of generating a incorrect text from the content of an image when replacing the synonym for a noun word. In this study, we propose a text augmentation method to replace words using word hierarchy information for noun words. Additionally, we performed experiments using MSCOCO data in order to evaluate the performance of the proposed methodology.

• Trans-
• /
• v.13
• /
• pp.141-169
• /
• 2022

Analysis of narrative texts has been regarded as academically and practically important, and has been made from various perspectives and methods. In this paper, the computational narrative analysis methodology from the perspective of information processing was examined. From the point of view of information processing, the creation and acceptance of narrative is a bidirectional coding process mediated by narrative text, and narrative text can be said to be a multi-layered structured code. In this paper, four methodologies that share this point of view - character network analysis, text mining and sentiment analysis, continuity analysis of event composition, and knowledge analysis of narrative agents - were examined together with cases. Through this, the mechanism and possibility of computational methodology in narrative analysis were confirmed. In conclusion, the significance and side effects of computational narrative analysis were examined, and the necessity of designing a human-computer collaboration model based on the consilience of the humanities and science/technology was discussed. Based on this model, it was argued that aesthetically creative, ethically good, politically progressive, and cognitively sophisticated narratives could be made more effectively.

• Journal of the Korea Society of Computer and Information
• /
• v.14 no.6
• /
• pp.1-10
• /
• 2009

Flash memory has the advantages of nonvolatile, low power consumption, light weight, and high endurance. This enables the flash memory to be utilized as a storage of mobile computing device such as PMP(Portable Multimedia Player). Potable device with a mass flash memory can store various multimedia data such as video, audio, or image. Typical index systems for mobile computer are inefficient to search a form of text like lyric or title. In this paper, we propose a new text index system, named EMTEX(Embedded Text Index). EMTEX has the following salient features. First, it uses a compression algorithm for embedded system. Second, if a new insert or delete operation is executed on the base table. EMTEX updates the text index immediately. Third, EMTEX considers the characteristics of flash memory to design insert, delete, and rebuild operations on the text index. Finally, EMTEX is executed as an upper layer of DBMS. Therefore, it is independent of the underlying DBMS. We evaluate the performance of EMTEX. The Experiment results show that EMTEX can outperform th conventional index systems such as Oracle Text and FT3.

• The Journal of the Convergence on Culture Technology
• /
• v.9 no.4
• /
• pp.185-193
• /
• 2023

The purpose of the study was to analyze the narrative meaning of the peritext in the picture-book of 『The Legend of Pat-bing-su』. For the purpose, based on the narrative components proposed by Nam and Kim, the narrative meanings of the peritext were analyzed. As the results, the peritexts of 『The Legend of Red Pat-bing-su』 include basic information of the title, author's name, and publication information, and physical elements of hard cover binding with matte rectangular paper that matches the narrative, which support prior understanding of the narratives. In addition, the peritext components such as covers, endpapers, title page, and copyright page lead readers to predict or expand narratives components to predict, expand, or transform the narrative, and provide additional information for understanding plots or genres.

• Journal of Korean Society of Archives and Records Management
• /
• v.22 no.2
• /
• pp.117-133
• /
• 2022

The archival content listed arbitrarily makes it difficult for users to efficiently access the records of major economic policies, especially given that they use it without understanding the required period and context. Using the text mining techniques in the 30-year economic policy direction from 1991 to 2021, this paper derives economic-related keywords and changes that the government mainly dealt with. It collects and preprocesses major economic policies' background, main content, and body text and conducts text frequency, term frequency-inverse document frequency (TF-IDF), network, and time series analyses. Based on these analyses, the following words are recorded in order of frequency: "job(일자리)," "competitive(경쟁력)," and "restructuring(구조조정)." In addition, the relative ratio of "job (일자리)," "real estate(부동산)," and "corporation(기업)," by year was analyzed in terms of chronological order while presenting major keywords mentioned by each government. Based on the results, this study presents implications for developing and broadening the area of archival information services related to economic policies.

• The Korean Journal of Applied Statistics
• /
• v.36 no.6
• /
• pp.501-514
• /
• 2023

When analyzing high dimensional data such as text data, if we input all the variables as explanatory variables, statistical learning procedures may suffer from over-fitting problems. Furthermore, computational efficiency can deteriorate with a large number of variables. Dimensionality reduction techniques such as feature selection or feature extraction are useful for dealing with these problems. The sparse principal component analysis (SPCA) is one of the regularized least squares methods which employs an elastic net-type objective function. The SPCA can be used to remove insignificant principal components and identify important variables from noisy observations. In this study, we propose a dimension reduction procedure for text data based on the SPCA. Applying the proposed procedure to real data, we find that the reduced feature set maintains sufficient information in text data while the size of the feature set is reduced by removing redundant variables. As a result, the proposed procedure can improve classification accuracy and computational efficiency, especially for some classifiers such as the k-nearest neighbors algorithm.

• Electronics and Telecommunications Trends
• /
• v.9 no.3
• /
• pp.97-102
• /
• 1994

상호대화형 오디오비주얼 서비스는 텍스트, 도형, 사진, 오디오, 비디오 등과 같은 다양한 형태의 표현 요소로 구성되는 입출력 정보를 사용자의 단말이나 워크스테이션에 제공하는 서비스이다. 이러한 기능의 범위는 간단한 검색에서부터 상호대화적인 문의, 구성요소들의 재배치, 그들 요소들의 수정등의 서비스를 사용자에게 제공 할 수 있다. 이와 관련하여 ITU-T SG8/Q.11에서는 AVI 서비스를 위해 요구되는, 시스템, 데이터 교환형식, 그리고 프로토콜과 같은 일련의 기술적 사항을 표준화하는 작업을 하고 있다. 본고에서는 AVI 서비스의 기술적인 사항에 대하여 논하고, 현재 진행되고 있는 표준화 동향에 대하여 알아본다.

• Proceedings of the Korean Society for Bioinformatics Conference
• /
• 2003.10a
• /
• pp.73-78
• /
• 2003

유전 연구를 통해 밝혀지고 있는 단백질은 각각의 기능적 특성을 가지고 서로 영향을 주고받으며 상호 작용한다. 단백질의 기능적 특성은 생물체에서는 단백질이 나타내는 기능으로 단백질 이름은 이들 단백질의 기능을 정확히 나타낼 수 있도록 붙여진다. 기능적 특성에 의해 명명된 단백질은 단백질을 구성하는 단어도 단백질과 유사한 기능 특성을 가질 가능성이 높다. 이는 텍스트 기반의 연구에서 단어가 가지는 중요성에서 비롯된다. 본 논문에서는 단백질을 구성하는 단어들을 단백질의 기능적 특성으로 분류하고, 이 기능분포에 의해서 단백질의 기능을 역으로 예측하고 판단하고자 하였다.

• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 1995.04a
• /
• pp.396-400
• /
• 1995

문화예술 정보는 텍스트, 이미지, 동화상등의 다양한 멀티미디어 데이타로 구성되어 있다. 이를 효과적으로 검색하기 위해서 노드와 링크로 구성된 하이퍼미디어를 사용하였다. 지금까지는 문화예술 정보중 연극 정보와 문화재 정보에 대하여 프로토타입을 구축하였으며, 계속하여 문화예술 전분야로 확대해 나갈 것이다. 연극정보를 검색하기 위해서는 데이타베이스 검색과 키워드 검색을 이용할 수 있으며, 최종적인 검색 결과는 분산하이퍼미디어 시스템인 Mosaic을 수정하여 이용하였다.

• Proceedings of the Korean Society for Information Management Conference
• /
• 1994.12a
• /
• pp.11-14
• /
• 1994

문화예술 정보는 텍스트, 이미지, 동화상등의 다양한 멀티미디어 데이타로 구성되어 있다. 이를 효과적으로 검색하기 위해서 노드와 링크로 구성된 하이퍼미디어를 사용하였다. 지금까지는 우선적으로 문화예술 정보중 연극 정보와 문화재 정보에 대하여 프로토타입을 구축하였으며, 계속하여 문화예술 전분야로 확대해 나갈 것이다. 연극정보를 검색하기 위해서는 데이타베이스 검색과 키워드 검색을 이용할 수 있으며, 최종적인 검색 결과는 하이퍼미디어 뷰어 (Hypermedia Viewer)인 Mosaic를 이용하였다.