한국어 데이터를 활용한 data augmentation

서가은(Gaeun Seo),오하영(Hayoung Oh) 한국정보통신학회 2023 한국정보통신학회논문지 Vol.27 No.4

오늘날 자연어 처리 분야의 발달에 따라 text 데이터 증강의 중요성이 강조되고 있다. 왜냐하면 모델의 정확도를 향상시키고 싶을 때, 자원이 한정된 현실에서 데이터 증강은 필수적이기 때문이다. 그러나 text 데이터 증강은 단어의 위치를 바꾸고 단어를 제거하는 등의 데이터를 변형하는 과정에서 원래의 뜻과 달라질 위험이 있다. 또한, 구어체로 이루어진 데이터의 경우, 문법에 맞지 않는 데이터가 많기 때문에 학습이 끝난 후 모델이 표준어가 아닌 문장을 입력으로 받았을 때 문장을 이해하지 못하는 경우가 있다. 따라서 본 논문에서는 이러한 문제점들을 해결하기 위해, 문장 데이터의 라벨을 보존하며 데이터를 증강하고, 문법에 맞지 않는 문장을 원본 데이터에 추가하는 증강 방법을 제안한다. With the development of natural language processing today, the importance of text data augmentation is being emphasized. This is because when you want to improve the accuracy of the model, data augmentation is essential in a reality with limited resources. However, text data augmentation risks being different from the original meaning in the process of transforming data, such as changing the position of words and removing words. In addition, in the case of colloquial data, there are many non-grammatical data, so after learning, the model may not understand the sentence when it receives a sentence that is not a standard language as an input. Therefore, in order to solve these problems, this paper proposes an augmented method that preserves the label of sentence data, enhances data, and adds non-grammatical sentences to the original data.

OLED의 광 효율 향상 기술: 랜덤 나노 외부 광 추출 복합 층 제작

박영욱,최근수,장은비,서가은 한국반도체디스플레이기술학회 2022 반도체디스플레이기술학회지 Vol.21 No.3

The light extraction technology for improving the light efficiency of OLEDs is the core technology for extracting the light inside the OLEDs to the outside. This study demonstrates a simple method to generate random nanostructures (RNSs) containing high refractive index nanoparticles to improve light extraction and viewing angle characteristics. A simple dry low-temperature process makes the nanostructured scattering layer on the polymer resin widely used in the industry. The scattering layer has the shape of randomly distributed nanorods. To control optical properties, we focused on changing the shape and density of RNSs and adjusting the concentration of high refractive index nanoparticles. As a result, the film of the present invention exhibits a perpendicular transmittance of 85% at a wavelength of 550 nm. This film was used as a scattering layer to reduce substrate mode loss and improve EL efficiency in OLEDs.

관절염을 가진 대상자의 가족지지, 자가 간호 및 삶의 질과의 관계

선은경,박소연,박일지,서가은,손용숙,손인옥,용진선 가톨릭대학교 간호대학 호스피스 교육연구소 2006 호스피스논집 Vol.10 No.-

Purpose : The purpose of this study was to investigate relationships among family support, self-care, and quality of life of patients with arthritis through a survey. That will be basic data to plan nursing intervention for promoting quality of life for patients with arthritis. Method : The convenient sample of this study consisted of 120 patients with arthritis and data were collected using questionnaires during August 2006 from two OPDs of rheumatoid centers in C University hospitals. The study instruments were used to measure Self Care (So, 1992), Family Support (Kang, 1984), and Quality of Life (Ro, 1988). The data were analyzed using ANOVA, Duncan test, and Pearson correlation. Result : The mean scores were 3.56 for family support, 4.16 for self-care, and 3.19 for quality of life. The level of self-care was significantly higher in those men, having an occupation, higher income, and living with family members. The level of family support was significantly higher in those men, having an occupation, living with family members, and having a helper. The level of quality of life was significantly higher for those men, having a higher income, and shorter length of illness. Three relationships were found to be positively correlated: 1) between the levels of family support and self-care (r=.53), 2) between the levels of self-care and quality of life (r=.55), and 3) between the levels of family support and quality of life (r=.54). Conclusion: Therefore, a nursing intervention program needs to be developed to promote family support, and self-care for quality of life for patients with arthritis.

RIE 공정을 이용한 유기발광다이오드의 광 산란층 제작

박영욱,장은비,배은정,최근수,서가은,장승미 한국반도체디스플레이기술학회 2022 반도체디스플레이기술학회지 Vol.21 No.1

Since the organic light-emitting diodes (OLEDs) have been widely investigated as next-generation displays, it has been successfully commercialized as a flexible and roll-able display. However, there is still wide room and demand to improve the device characteristics such as power efficiency and lifetime. To solve this issue, there has been a wide research effort, and among them, the internal and the external light extraction techniques have been attracted in this research field by its fascinating characteristic of material independence. In this study, a micro-nano composite structured external light extraction layer was demonstrated. A reactive ion etching (RIE) process was performed on the surfaces of hexagonally packed hemisphere micro-lens-array (MLA) and randomly distributed sphere diffusing films to form micro-nano composite structures. Random nanostructures of different sizes were fabricated by controlling the processing time of the O2 / CHF3 plasma. The fabricated device using a micro-nano composite external light extraction layer showed 1.38X improved external quantum efficiency compared to the reference device. The results prove that the external light extraction efficiency is improved by applying the micro-nano composite structure on conventional MLA fabricated through a simple process.

환자 IQR 이상치와 상관계수 기반의 머신러닝 모델을 이용한 당뇨병 예측 메커니즘

정주호,이나은,김수민,서가은,오하영,Jung, Juho,Lee, Naeun,Kim, Sumin,Seo, Gaeun,Oh, Hayoung 한국정보통신학회 2021 한국정보통신학회논문지 Vol.25 No.10

최근 전 세계적으로 당뇨병 유발률이 증가함에 따라 다양한 머신러닝과 딥러닝 기술을 통해 당뇨병을 예측하려고 는 연구가 이어지고 있다. 본 연구에서는 독일의 Frankfurt Hospital 데이터로 머신러닝 기법을 활용하여 당뇨병을 예측하는 모델을 제시한다. IQR(Interquartile Range) 기법을 이용한 이상치 처리와 피어슨 상관관계 분석을 적용하고 Decision Tree, Random Forest, Knn, SVM, 앙상블 기법인 XGBoost, Voting, Stacking로 모델별 당뇨병 예측 성능을 비교한다. 연구를 진행한 결과 Stacking ensemble 기법의 정확도가 98.75%로 가장 뛰어난 성능을 보였다. 따라서 해당 모델을 이용하여 현대 사회에 만연한 당뇨병을 정확히 예측하고 예방할 수 있다는 점에서 본 연구는 의의가 있다. With the recent increase in diabetes incidence worldwide, research has been conducted to predict diabetes through various machine learning and deep learning technologies. In this work, we present a model for predicting diabetes using machine learning techniques with German Frankfurt Hospital data. We apply outlier handling using Interquartile Range (IQR) techniques and Pearson correlation and compare model-specific diabetes prediction performance with Decision Tree, Random Forest, Knn (k-nearest neighbor), SVM (support vector machine), Bayesian Network, ensemble techniques XGBoost, Voting, and Stacking. As a result of the study, the XGBoost technique showed the best performance with 97% accuracy on top of the various scenarios. Therefore, this study is meaningful in that the model can be used to accurately predict and prevent diabetes prevalent in modern society.

Development of Monoclonal Antibodies for Diagnosis of Plasmodium vivax

Nguyen Thi Phuong Linh,박현,이진영,Dong-Xu Liu,서가은,손혜진,한진희,한은택,신호준,여선주 대한기생충학ㆍ열대의학회 2017 The Korean Journal of Parasitology Vol.55 No.6

Plasmodium lactate dehydrogenase (pLDH) is a strong target antigen for the determination of infection with Plasmodium species specifically. However, a more effective antibody is needed because of the low sensitivity of the current antibody in many immunological diagnostic assays. In this study, recombinant Plasmodium vivax LDH (PvLDH) was experimentally constructed and expressed as a native antigen to develop an effective P. vivax-specific monoclonal antibody (mAb). Two mAbs (2CF5 and 1G10) were tested using ELISA and immunofluorescence assays (IFA), as both demonstrated reactivity against pLDH antigen. Of the 2 antibodies, 2CF5 was not able to detect P. falciparum, suggesting that it might possess P. vivax-specificity. The detection limit for a pair of 2 mAbs-linked sandwich ELISA was 31.3 ng/ml of the recombinant antigen. The P. vivax-specific performance of mAbs-linked ELISA was confirmed by in vitro-cultured P. falciparum and P. vivax-infected patient blood samples. In conclusion, the 2 new antibodies possessed the potential to detect P. vivax and will be useful in immunoassay.