• 한국융합학회논문지
• /
• 제11권9호
• /
• pp.21-29
• /
• 2020

최근 디지털 사회의 도래로 다양한 데이터가 폭발적으로 증가하고, 그중 문헌 내 주제어를 도출하는 토픽 모델링에 관한 연구가 활발히 진행되고 있다. 본 논문의 연구목표는 토픽 모델링 방법 중 하나인 DTM(Dynamic Topic Model) 모델을 적용해 D.N.A.(Data, Network, A.I) 분야에 대한 연구동향을 탐색하는데 있다. 실험 데이터는 최근 6년간(2015~2020) ICT(Information and Communication Technology) 분야 중 기술대분류가 SW·AI에 해당하는 연구과제 1,519개 사업에 대해 DTM 모델을 적용하였다. 실험결과로, D.N.A. 분야의 기술 키워드 Big data, Cloud, Artificial Intelligence와 확장된 의미의 기술 키워드 Unstructured, Edge Computing, Learning, Recognition 등이 매년 연구에 표출되었으며, 해당 키워드 들이 특정 연구과제에 종속되지 않고 다른 연구과제에서도 포괄적으로 연구되고 있음을 확인하였다. 끝으로 본 논문의 연구결과는 향후 D.N.A. 분야에 대한 정책기획·과제기획 등 연구개발 기획 과정과 기업의 기술 확보전략·마케팅 전략 등 다양한 곳에 활용될 수 있을 것으로 기대한다.

• Journal of Information Processing Systems
• /
• 제15권6호
• /
• pp.1306-1325
• /
• 2019

As the world is moving towards digitization, data is generated from various sources at a faster rate. It is getting humungous and is termed as big data. The financial sector is one domain which needs to leverage the big data being generated to identify financial risks, fraudulent activities, and so on. The design of predictive models for such financial big data is imperative for maintaining the health of the country's economics. Financial data has many features such as transaction history, repayment data, purchase data, investment data, and so on. The main problem in predictive algorithm is finding the right subset of representative features from which the predictive model can be constructed for a particular task. This paper proposes a correlation-based method using submodular optimization for selecting the optimum number of features and thereby, reducing the dimensions of the data for faster and better prediction. The important proposition is that the optimal feature subset should contain features having high correlation with the class label, but should not correlate with each other in the subset. Experiments are conducted to understand the effect of the various subsets on different classification algorithms for loan data. The IBM Bluemix BigData platform is used for experimentation along with the Spark notebook. The results indicate that the proposed approach achieves considerable accuracy with optimal subsets in significantly less execution time. The algorithm is also compared with the existing feature selection and extraction algorithms.

• Mycobiology
• /
• 제49권6호
• /
• pp.582-588
• /
• 2021

The interaction of mating pheromone and pheromone receptor from the B mating-type locus is the first step in the activation of the mushroom mating signal transduction pathway. The B mating-type locus of Lentinula edodes is composed of Bα and Bβ subloci, each of which contains genes for mating pheromone and pheromone receptor. Allelic variations in both subloci generate multiple B mating-types through which L. edodes maintains genetic diversity. In addition to the B mating-type locus, our genomic sequence analysis revealed the presence of a novel chromosomal locus 43.3 kb away from the B mating-type locus, containing genes for a pair of mating pheromones (PHBN1 and PHBN2) and a pheromone receptor (RCBN). The new locus (Bα-N) was homologous to the Bα sublocus, but unlike the multiallelic Bα sublocus, it was highly conserved across the wild and cultivated strains. The interactions of RcbN with various mating pheromones from the B and Bα-N mating-type loci were investigated using yeast model that replaced endogenous yeast mating pheromone receptor STE2 with RCBN. The yeast mating signal transduction pathway was only activated in the presence of PHBN1 or PHBN2 in the RcbN producing yeast, indicating that RcbN interacts with self-pheromones (PHBN1 and PHBN2), not with pheromones from the B mating-type locus. The biological function of the Bα-N locus was suggested to control the expression of A mating-type genes, as evidenced by the increased expression of two A-genes HD1 and HD2 upon the treatment of synthetic PHBN1 and PHBN2 peptides to the monokaryotic strain of L. edodes.

• Journal of the Korean Data and Information Science Society
• /
• 제26권6호
• /
• pp.1259-1269
• /
• 2015

최근 빅 데이터는 학계에서 키워드로 자리매김을 하고 있다. 빅 데이터의 유용성은 학계뿐만 아니라 정부, 지자체 그리고 기업체까지 파급되고 있고, 빅 데이터 속에서 유용한 정보를 도출해 내기 위해 노력하고 있다. 본 연구에서는 영화에 대한 리뷰를 가지고 텍스트 마이닝 (text mining)을 이용한 빅 데이터 분석을 수행한다. 본 연구의 목적은 포털 사이트 'D'사와 영화진흥위원회의 영화에 대한 리뷰 데이터, 그리고 고객들의 평점평균 (score)과 스크린 수 (screen number)를 설명변수로 사용하고, 영화 흥행 여부를 종속변수로 하여 로지스틱 회귀분석을 통한 영화 흥행 예측 모형을 제안하는 것이다. 분석결과, 본 연구에서 제안한 예측모형의 정분류율은 95.74%로 얻어졌다.

• 한국ITS학회 논문지
• /
• 제22권6호
• /
• pp.157-167
• /
• 2023

본 연구에서는 분산 컴퓨팅 및 개별 디바이스 활용을 통해 개인 정보 보호에 특화된 학습방법인 연합학습 방법론을 기반으로, 모바일 내비게이션 애플리케이션에서 수집된 대규모의 운전자 데이터를 이용하여 경로 선택 예측 모델을 수립하는 방법에 대해 고찰한다. 경로 선택 모델링에서 활용될 수 있는 운전자 데이터의 전처리 및 분석 방법을 수립하고, 서포트벡터머신(SVM) 및 다층 퍼셉트론(MLP)과 같이 기존에 널리 활용되는 학습 방법과 연합학습 방법의 성능과 특성을 비교한다. 분석 결과 연합학습을 통한 모델 성능은 중앙 서버 기반의 모델과의 비교에서 예측 정확도 측면의 차이가 거의 없는 것으로 나타났으나, 개별 데이터가 충분히 확보되는 경우 연합학습 모델과 같은 개인화 모델의 성능이 개선될 수 있다는 점을 확인하였다. 연합학습 모델은 본 연구의 경로 선택 모델링 사례와 같이 모빌리티 부문의 데이터 프라이버시 문제가 중요한 분야에서 대규모 데이터 처리를 필요로 하는 경우에 그 활용 가치가 매우 높을 것으로 기대된다.

• 한국인공지능학회지
• /
• 제6권1호
• /
• pp.16-20
• /
• 2018

The development of modern information technology has increased the amount of big data about patients' information and diseases. In this study, we developed a prediction model of diabetes using the health examination data provided by the public data portal in 2016. In addition, we graphically visualized diabetes incidence by sex, age, residence area, and income level. As a result, the incidence of diabetes was different in each residence area and income level, and the probability of accurately predicting male and female was about 65%. In addition, it can be confirmed that the influence of X on male and Y on female is highly to affect diabetes. This predictive model can be used to predict the high-risk patients and low-risk patients of diabetes and to alarm the serious patients, thereby dramatically improving the re-admission rate. Ultimately it will be possible to contribute to improve public health and reduce chronic disease management cost by continuous target selection and management.

• Computers and Concrete
• /
• 제24권6호
• /
• pp.571-578
• /
• 2019

This study aimed to investigate the permeable nature of pervious concretes (PC) through the moving particle simulation (MPS) method. In the simulation, the complex structure of a pervious concrete was virtually demonstrated as a lattice model (LM) of spherical beads, where the test of permeability was conducted. Results of the simulation were compared with the experimental ones for validation. As a result, MPS results showed the permeability index of the LM as almost twice as big as the actual PCs. A proposed virtual model was created to prevent the stuck of water flow in the MPS simulation of PC or LM. Successful simulation results were demonstrated with the model.

• 천문학회보
• /
• 제37권2호
• /
• pp.111.2-111.2
• /
• 2012

We observed an Ellerman bomb(EB) and its associated surge using the Fast Imaging Solar Spectrograph(FISS) and the broadband TiO filter of the 1.6 meter New Solar Telescope at Big Bear Solar Observatory. As is well-known, the EB appears as a feature that is very bright at the far wings of the H alpha line. The lambdameter method applied to these wings indicates that the EB is blue-shifted up to 6km/s in velocity. In the photospheric level below the EB, we see rapidly growing "granule-like" feature. The transverse velocity of the dark lane at the edge of the "granule" increased with time as reached a peak of 6km/s, at the time of the EB's occurrence. The surge was seen in absorption and varied rapidly both in the H alpha and the Ca II 8542 line. It originated from the Ellerman bomb, and was impulsively accelerated to 20km/s toward us(blueshift). Then the velocity of the surge gradually changed from blueshift of 20km/s to redshift of 40km/s. By adopting the cloud model, we estimated the temperature of the surge material at about 27000K and the non-thermal velocity at about 10km/s. Our results shed light on the conventional idea that an EB results from the magnetic reconnection of an emerging flux tube and pre-existing field line.

• 스마트미디어저널
• /
• 제6권4호
• /
• pp.88-93
• /
• 2017

"지하국대적퇴치설화"의 서사는 세계에 분포하는 설화로 나라마다 이야기와 구조가 크게 다르지 않다. 이런 익숙한 서사구조는 다른 문화를 가진 곳에서도 큰 거부감 없이 받아들여질 수 있다. 선행연구에 따라 설화에서 추출한 요소를 가지고 스토리텔링을 진행할 경우에는 설화의 개연성과 보편성이 확보된 콘텐츠 창작이 가능함으로 설화의 요소와 구조를 이용하여 그림책 스토리텔링을 진행하였다. 설화를 콘텐츠화 하는 방식 중 하나를 새로쓰기라고 하는데 본 논문에서는 설화의 구성요소를 분석한 후 그림책의 구성요소를 문학적 관점에서 주제, 플롯, 등장인물, 배경으로 나누어 그림책 스토리텔링을 진행함으로 "지하국대적퇴치설화"의 새로쓰기 방법을 제시하였다.

• International Journal of Internet, Broadcasting and Communication
• /
• 제14권4호
• /
• pp.41-46
• /
• 2022

The purpose of this study is to confirm and analyze the impact on consumers through big data keyword analysis on weak food. For data collection, web documents, blogs, news, cafes, intellectuals, academic information, and Google Web, news, and Facebook provided by Naver and Daum were used as analysis targets. The data analysis period was set from January 2018 to December 2021. For data collection and analysis, the frequency and matrix of keywords were extracted through Textom, a social matrix site, and the relationship and connection centrality between keywords were analyzed and visualized using the Netdraw function among UCINET6 programs. In addition, CONCOR analysis was conducted to derive clusters for similar keywords. As a result of analyzing yakseon food with keywords, a total of 35,985 cases of collected data were derived. Through this, it was confirmed that medicinal food affects consumers. Furthermore, if a business model is created and developed through yakseon food, it will be possible to lead the popularization of yakseon food.