http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
PIAO XIANGHUA,YIN HELIN,구영현(Yeong Hyeon Gu),유성준(Seong Joon Yoo) 한국방송·미디어공학회 2020 한국방송공학회 학술발표대회 논문집 Vol.2020 No.7
국가기후기술정보시스템은 국내 환경기술과 국외의 수요기술 정보를 제공하는 검색 시스템이다. 그러나 기존의 시스템은 유사한 뜻을 가진 단일 단어와 복수 단어들을 모두 식별하지 못하기에 유의어를 입력했을 경우 검색결과가 다르다. 이런 문제점을 해결하기 위해 본 연구에서는 유의어 사전을 기반으로한 환경기술 검색 시스템을 제안한다. 이 시스템은 Word2vec 모델과 HDBSCAN(Hierarchical Density-Based Spatial Clustering of Application with Noise) 알고리즘을 이용해 유의어 사전을 구축한다. Word2vec 모델을 이용해 한국어와 영어 위키백과 코퍼스에 대해 형태소 분석을 진행한 후 단일 단어와 복수 단어를 포함한 단어를 추출하고 벡터화를 진행한다. 그 다음 HDBSCAN 알고리즘을 이용해 벡터화된 단어를 군집화 해주고 유의어를 추출한다. 기존의 Word2vec 모델이 모든 단어 간의 거리를 계산하고 유의어를 추출하는 과정과 대비하면 시간이 단축되는 역할을 한다. 추출한 유의어를 통합해 유의어 사전을 구축한다. 국가기후기술정보시스템에서 제공하는 국내외 기술정보, 기술정보 키워드와 구축한 유의어 사전을 Multi-filter를 제공하는 Elasticsearch에 적용해 최종적으로 유의어를 식별할 수 있는 환경기술 검색 시스템을 제안한다.
한국어 언어학적 특징 기반 자동화된 사실 문장과 의견 분류
김은비(Eun Bi Kim),PIAO XIANGHUA,손민주(Min ju Seon),유성준(Seong jun Yoo),구영현(Yeong Hyeon Gu),탁진영(Jinyoung Tak) 대한전자공학회 2023 대한전자공학회 학술대회 Vol.2023 No.6
The importance of distinguishing between facts and opinions in the field of journalism has been emphasized for the purpose of delivering fair and balanced reporting. However, in practice, there have been challenges in effectively differentiating between facts and opinions. This study proposes automated model that can distinguish factual information and opinions based on linguistic features using the text data from Korean language textbooks for 2nd to 3rd grade in secondary education. The experiment using KcBERT which is a Korean deep learning language model on a dataset of 941 sentences showed a performance result of 87.30% accuracy.
WideResNet 및 이미지 보간 기법을 적용한 UTRAD 모델 연구
권민정(Min Jung Kwon),PIAO XIANGHUA,PIAO ZHEGAO,구영현(Yeong Hyeon Gu) 대한전자공학회 2023 대한전자공학회 학술대회 Vol.2023 No.11
Deep learning models have shown remarkable performance in the field of anomaly detection, but model training is often influenced by the issue of data imbalance. In this paper, we conducted performance enhancement experiments using the UTRAD model, which addresses the aforementioned problem. The experimental results demonstrated a 1.61% improvement in Image-AUROC by applying the WideResNet backbone and image interpolation techniques to UTRAD.
박성열(Seong Yeol Park),PIAO XIANGHUA,JIN DONG,YIN HELIN,구영현(Yeong Hyeon Gu) 대한전자공학회 2023 대한전자공학회 학술대회 Vol.2023 No.6
As the proportion of aquaculture in the fisheries industry increases, it becomes increasingly important to address the problems that arise. To prevent the rapid spread of diseases in fish farms, this paper proposes a deep learning-based image classification model. The performance of VGG16, ResNet50, Xception, MViTv2, DaViT, and CoAtNet was compared for the classification of fish diseases such as bleeding, defects, and necrosis. Precision, recall, F1 score, and accuracy were used as performance metrics. Our experimental findings reveal that the CoAtNet model exhibits the highest recall and accuracy rates, both reaching 0.6410.
Anomaly Prediction Model Using Warning Signs
Yoojin Ha,Won Hee Chung,Xianghua Piao,Yeong Hyeon Gu,SEOGBONG JEON,Seong Joon Yoo 한국차세대컴퓨팅학회 2021 한국차세대컴퓨팅학회 학술대회 Vol.2021 No.11
Generators continue to deteriorate in performance due to aging and result in increased failure rates and reduced reliability. Therefore, studies are being conducted on anomaly prediction models for generator engines to prevent potential accidents during operation. However, there are problems in designing the models due to class imbalance and manual input of maintenance history. This study labels data from the time an anomaly occurs up to 60 minutes before the occurrence as anomalies to solve these problems. Data from the time an anomaly occurs up to 30 minutes before the occurrence were also added as derived variables to reflect the warning signs of anomalies in model training. The anomaly prediction models were created using engine log and maintenance history data and applying Random Forest(RF), eXtreme Gradient Boosting(XGB), Linear Support Vector Classifier(LSVC), and Deep Neural Networks(DNN) algorithms. The performance of the models was evaluated by F1-Score and Recall. XGB showed excellent performance in terms of F1-Score, and DNN in terms of Recall. As a result of comparing the F1-Scores to sort the optimal model for each system, XGB was optimal for systems 1, 2, and 4, and RF was optimal for systems 3 and 5. System 5 showed excellent performance when only the derived variable condition was applied, and the other systems showed excellent performance when applying the derived variable and labeling.