• Proceedings of the Korea Information Processing Society Conference
• /
• 2019.10a
• /
• pp.820-822
• /
• 2019

개체명이란, 문서에서 특정한 의미를 가지고 있는 단어나 어구를 뜻하는 말로 사람, 기관명, 지역명, 날짜, 시간 등이 있으며 이 개체명을 찾아서 해당하는 의미의 범주를 결정하는 것을 개체명 인식이라고 한다. 본 논문에서는 BERT(Bidirectional Encoder Representations from Transformers) 활용한 한국어 개체명 인식기를 제안한다. 제안하는 모델은 기 학습된 BERT 모델을 활용함으로써 성능을 극대화하여, 최종 F1-Score 는 90.62 를 달성하였고, Bi-LSTM-Attention-CRF 모델에 비해 매우 뛰어난 결과를 보였다.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• fall
• /
• pp.314-317
• /
• 2021

하이라이트 영상은 원본 영상의 중요한 장면들을 짧은 시간 안에 감상할 수 있게 도와준다. 특히나 경기 시간 긴 축구나 야구 그리고 e-스포츠의 시청자들에게 있어, 하이라이트 영상의 효용성은 더욱 증가한다. 하이라이트 영상 추출의 자동화로 방송사나 온라인 플랫폼은 비용 절감과 시간 절약의 이점을 얻을 수 있다. 따라서 본 논문에서는 스포츠 영상에서 자동으로 하이라이트 구간을 추출하는 모델을 제안한다. 제안하는 모델은 멀티 헤드 어텐션 매커니즘과 LSTM 네트워크의 결합으로 구성된다. 해당 매커니즘의 여러 헤드를 통해 어텐션을 다양한 관점에서 진행한다. 이로 인해 영상의 전체적인 맥락과 장면 간의 유기적 관계를 다양한 관점에서 파악할 수 있다. 또한 오디오와 이미지 정보를 함께 이용하여 모델을 학습한다. 학습한 모델의 평가는 e-스포츠 경기 영상을 이용하여 평가한다.

• Annual Conference on Human and Language Technology
• /
• 2021.10a
• /
• pp.329-334
• /
• 2021

한국어에서는 문장 내의 주어나 목적어가 자주 생략된다. 자연어 처리에서 이러한 문장을 그대로 사용하는 것은 정보 부족으로 인한 문제 난이도 상승으로 귀결된다. 생략복원은 텍스트에서 생략된 부분을 이전 문구에서 찾아서 복원해 주는 기술이며, 본 논문은 생략된 주어를 복원하는 방법에 대한 연구이다. 본 논문에서는 기존에 생략복원에 사용되지 않았던 다양한 입력 형태를 시도한다. 또한, 출력 레이어로는 finetuning layer(Linear, Bi-LSTM, MultiHeadAttention)와 생략복원 태스크 형태(BIO tagging, span prediction)의 다양한 조합을 실험한다. 국립국어원 무형 대용어 복원 말뭉치를 기반으로 생략복원이 불필요한 네거티브 샘플을 추가하여 ELECTRA 기반의 딥러닝 생략복원 모델을 학습시키고, 생략복원에 최적화된 조합을 검토한다.

• The Journal of the Acoustical Society of Korea
• /
• v.39 no.6
• /
• pp.592-599
• /
• 2020

This paper proposes a Convolutional Recurrent Neural Net (CRNN) structure that can simultaneously reflect both static and dynamic characteristics of seismic waveforms for various earthquake events classification. Addressing various earthquake events, including not only micro-earthquakes and artificial-earthquakes but also macro-earthquakes, requires both effective feature extraction and a classifier that can discriminate seismic waveform under noisy environment. First, we extract the static characteristics of seismic waveform through an attention-based convolution layer. Then, the extracted feature-map is sequentially injected as input to a multi-input single-output Long Short-Term Memory (LSTM) network structure to extract the dynamic characteristic for various seismic event classifications. Subsequently, we perform earthquake events classification through two fully connected layers and softmax function. Representative experimental results using domestic and foreign earthquake database show that the proposed model provides an effective structure for various earthquake events classification.

• Journal of Internet Computing and Services
• /
• v.22 no.5
• /
• pp.79-86
• /
• 2021

As the usefulness of big data analysis has been drawing attention, many studies in the business research area begin to use big data to predict firm performance. Previous studies mainly rely on data outside of the firm through news articles and social media platforms. The voices within the firm in the form of employee satisfaction or evaluation of the strength and weakness of the firm can potentially affect firm value. However, there is insufficient evidence that online employee reviews are valid to predict firm value because the data is relatively difficult to obtain. To fill this gap, from 2014 to 2019, we employed 97,216 reviews collected by JobPlanet, an online firm review website in Korea, and developed a machine learning-based predictive model. Among the proposed models, the LSTM-based model showed the highest accuracy at 73.2%, and the MAE showed the lowest error at 0.359. We expect that this study can be a useful case in the field of firm value prediction on domestic companies.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.25 no.1
• /
• pp.1-7
• /
• 2021

Investor mood from social media is gaining increasing attention for leading a price movement in stock market. Based on the behavioral finance theory, this study argues that sentiment extracted from social media using big data technique can predict a real-time (short-run) price momentum in Chinese stock market. Collecting Sina Weibo posts that related to COVID-19 using keyword method, a daily influential weighted sentiment factors is extracted from the sizable raw data of over 2 millions of posts. We examine one supervised and 4 unsupervised sentiment analysis model, and use the best performed word-frequency and BiLSTM mdoel. The test result shows a similar movement between stock price change and sentiment factor. It indicates that public mood extracted from social media can in some extent represent the investors' sentiment and make a difference in stock market fluctuation when people are concentrating on a special events that can cause effect on the stock market.

• International Journal of Computer Science & Network Security
• /
• v.22 no.10
• /
• pp.246-256
• /
• 2022

Recently, parents and teachers consider physical education as a minor subject for students in elementary and secondary schools. Physical education performance has become increasingly significant as parents and schools pay more attention to physical schooling. The sports mining with distribution analysis model considers different factors, including the games, comments, conversations, and connection made on numerous sports interests. Using different machine learning/deep learning approach, children's athletic and academic interests can be tracked over the course of their academic lives. There have been a number of studies that have focused on predicting the success of students in higher education. Sports interest prediction research at the secondary level is uncommon, but the secondary level is often used as a benchmark to describe students' educational development at higher levels. An Automated Student Interest Prediction on Sports Mining using DL Based Bi-directional Long Short-Term Memory model (BiLSTM) is presented in this article. Pre-processing of data, interest classification, and parameter tweaking are all the essential operations of the proposed model. Initially, data augmentation is used to expand the dataset's size. Secondly, a BiLSTM model is used to predict and classify user interests. Adagrad optimizer is employed for hyperparameter optimization. In order to test the model's performance, a dataset is used and the results are analysed using precision, recall, accuracy and F-measure. The proposed model achieved 95% accuracy on 400^th instances, where the existing techniques achieved 93.20% accuracy for the same. The proposed model achieved 95% of accuracy and precision for 60%-40% data, where the existing models achieved 93% for accuracy and precision.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.46 no.3
• /
• pp.123-138
• /
• 2023

The Fourth Industrial Revolution and sensor technology have led to increased utilization of sensor data. In our modern society, data complexity is rising, and the extraction of valuable information has become crucial with the rapid changes in information technology (IT). Recurrent neural networks (RNN) and long short-term memory (LSTM) models have shown remarkable performance in natural language processing (NLP) and time series prediction. Consequently, there is a strong expectation that models excelling in NLP will also excel in time series prediction. However, current research on Transformer models for time series prediction remains limited. Traditional RNN and LSTM models have demonstrated superior performance compared to Transformers in big data analysis. Nevertheless, with continuous advancements in Transformer models, such as GPT-2 (Generative Pre-trained Transformer 2) and ProphetNet, they have gained attention in the field of time series prediction. This study aims to evaluate the classification performance and interval prediction of remaining useful life (RUL) using an advanced Transformer model. The performance of each model will be utilized to establish a health index (HI) for cutting blades, enabling real-time monitoring of machine health. The results are expected to provide valuable insights for machine monitoring, evaluation, and management, confirming the effectiveness of advanced Transformer models in time series analysis when applied in industrial settings.

• The Journal of the Korea Contents Association
• /
• v.18 no.11
• /
• pp.416-429
• /
• 2018

Typhoons and floods are natural disasters that occur frequently, and the damage resulting from these disasters must be in advance predicted to establish appropriate responses. Direct damages such as building collapse, human casualties, and loss of farms and fields have more attention from people than indirect damages such as increase of consumer prices. But indirect damages also need to be considered for living. The agricultural products are typical consumer items affected by typhoons and floods. Sudden, powerful typhoons are mostly accompanied by heavy rains and damage agricultural products; this increases the retail price of such products. This study analyzes the influence of natural disasters on the price of agricultural products by using a deep learning algorithm. We decided rice, onion, green onion, spinach, and zucchini as target agricultural products, and used data on variables that influence the price of agricultural products to create a model that predicts the price of agricultural products. The result shows that the model's accuracy was about 0.069 measured by RMSE, which means that it could explain the changes in agricultural product prices. The accurate prediction on the price of agricultural products can be utilized by the government to respond natural disasters by controling amount of supplying agricultural products.

• Knowledge Management Research
• /
• v.22 no.2
• /
• pp.19-32
• /
• 2021

Bitcoin, a representative cryptocurrency, is receiving a lot of attention around the world, and the price of Bitcoin shows high volatility. High volatility is a risk factor for investors and causes social problems caused by reckless investment. Since the price of Bitcoin responds quickly to changes in the world environment, we propose to predict the price volatility of Bitcoin by utilizing news information that provides a variety of information in real-time. In other words, positive news stimulates investor sentiment and negative news weakens investor sentiment. Therefore, in this study, sentiment information of news and deep learning were applied to predict the change in Bitcoin yield. A single predictive model of logit, artificial neural network, SVM, and LSTM was built, and an integrated model was proposed as a method to improve predictive performance. As a result of comparing the performance of the prediction model built on the historical price information and the prediction model reflecting the sentiment information of the news, it was found that the integrated model based on the sentiment information of the news was the best. This study will be able to prevent reckless investment and provide useful information to investors to make wise investments through a predictive model.