• Annual Conference on Human and Language Technology
• /
• 2020.10a
• /
• pp.60-63
• /
• 2020

기계 독해는 입력 받은 질문과 문단의 관계를 파악하여 알맞은 정답을 예측하는 자연어처리 태스크로 양질의 많은 데이터 셋을 필요로 한다. 기계 독해 학습 데이터 구축은 어려운 작업으로, 문서에서 등장하는 정답과 정답을 도출할 수 있는 질문을 수작업으로 만들어야 한다. 이러한 문제를 해결하기 위하여, 본 논문에서는 정답이 속한 문서로부터 질문을 자동으로 생성해주는 BERT 기반의 Sequence-to-sequence 모델을 이용한 한국어 질문 생성 모델을 제안한다. 또한 정답이 속한 문서와 질문의 언어가 같고 정답이 속한 문장의 주변 단어가 질문에 등장할 확률이 크다는 특성에 따라 BERT 기반의 Sequence-to-sequence 모델에 복사 메카니즘을 추가한다. 실험 결과, BERT + Transformer 디코더 모델의 성능이 기존 모델과 BERT + GRU 디코더 모델보다 좋았다.

• Journal of the Korean Society of Marine Environment & Safety
• /
• v.27 no.5
• /
• pp.574-583
• /
• 2021

Predicting shipping markets is an important issue. Such predictions form the basis for decisions on investment methods, fleet formation methods, freight rates, etc., which greatly affect the profits and survival of a company. To this end, in this study, we propose a shipping freight rate prediction model for container ships using gated recurrent units (GRUs) and long short-term memory structure. The target of our freight rate prediction is the China Container Freight Index (CCFI), and CCFI data from March 2003 to May 2020 were used for training. The CCFI after June 2020 was first predicted according to each model and then compared and analyzed with the actual CCFI. For the experimental model, a total of six models were designed according to the hyperparameter settings. Additionally, the ARIMA model was included in the experiment for performance comparison with the traditional analysis method. The optimal model was selected based on two evaluation methods. The first evaluation method selects the model with the smallest average value of the root mean square error (RMSE) obtained by repeating each model 10 times. The second method selects the model with the lowest RMSE in all experiments. The experimental results revealed not only the improved accuracy of the deep learning model compared to the traditional time series prediction model, ARIMA, but also the contribution in enhancing the risk management ability of freight fluctuations through deep learning models. On the contrary, in the event of sudden changes in freight owing to the effects of external factors such as the Covid-19 pandemic, the accuracy of the forecasting model reduced. The GRU1 model recorded the lowest RMSE (69.55, 49.35) in both evaluation methods, and it was selected as the optimal model.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2023.11a
• /
• pp.932-933
• /
• 2023

이 연구는 사람의 감정 변화를 건강하게 파악하고 분석하기 위해 시작되었다. Natural Language Processing(NLP)는 컴퓨터가 인간의 언어를 이해하기 위해 개발된 자연어 처리 기술이다. 본 논문에서는 이 기술을 이용하여 Text Mining을 통해 사용자가 작성한 일기에 담긴 감정을 분석하고 LSTM 모델과 GRU 모델을 비교군으로 두어 두 모델 중 감정 분석에 더 적합한 모델을 찾는 과정을 거쳤다. 이 과정을 정확도가 더 높은 LSTM 모델을 사용하여 감정 분석 결과를 분류하였다.

• Annual Conference on Human and Language Technology
• /
• 2018.10a
• /
• pp.593-595
• /
• 2018

자동 띄어쓰기는 띄어쓰기가 되어있지 않은 문장에 대하여 띄어쓰기를 해주거나, 문장에 있는 잘못된 띄어쓰기를 교정하는 것을 말한다. 기존의 자동 띄어쓰기 연구는 주로 모든 음절을 붙인 후 새로 띄어쓰기 태그를 입력하는 방법을 사용하여 사용자가 입력한 올바른 띄어쓰기 정보를 활용하지 못하였다. 따라서 본 논문에서는 모두 붙여 쓴 문장에 공백을 넣어주는 띄어쓰기 삽입 모델과 사용자의 입력 정보를 이용하여 문장의 띄어쓰기 오류를 교정해주는 오류교정 모델이 결합된 통합모델을 제안한다. 제안된 모델은 에러율 10%일 때 F1-score가 98.85%까지 향상되었다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2023.11a
• /
• pp.1106-1107
• /
• 2023

자율주행 선박의 기술 발전에 따라 적대적 공격에 대한 위험성이 대두되고 있다. 이를 해결하기 위해 본 연구는 다양한 신경망 모델을 활용하여 적대적 공격을 탐지하는 성능을 체계적으로 비교, 분석하였다. CNN, GRU, LSTM, VGG16 모델을 사용하여 실험을 진행하였고, 이 중 VGG16 모델이 가장 높은 탐지 성능을 보였다. 본 연구의 결과를 통해 자율주행 선박에 적용될 수 있는 보안모델 구축에 대한 신뢰성 있는 방향성을 제시하고자 한다.

• Journal of Korean Society on Water Environment
• /
• v.39 no.1
• /
• pp.46-60
• /
• 2023

Climate change causes fluctuations in water quality in the aquatic environment, which can cause changes in water circulation patterns and severe adverse effects on aquatic ecosystems in the future. Therefore, research is needed to predict and respond to water quality changes caused by climate change in advance. In this study, we tried to predict the dissolved oxygen (DO), chlorophyll-a, and turbidity of the Paldang reservoir for about two weeks using long short-term memory (LSTM) and gated recurrent units (GRU), which are deep learning algorithms based on recurrent neural networks. The model was built based on real-time water quality data and meteorological data. The observation period was set from July to September in the summer of 2021 (Period 1) and from March to May in the spring of 2022 (Period 2). We tried to select an algorithm with optimal predictive power for each water quality parameter. In addition, to improve the predictive power of the model, an important variable extraction technique using random forest was used to select only the important variables as input variables. In both Periods 1 and 2, the predictive power after extracting important variables was further improved. Except for DO in Period 2, GRU was selected as the best model in all water quality parameters. This methodology can be useful for preventive water quality management by identifying the variability of water quality in advance and predicting water quality in a short period.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.30 no.4
• /
• pp.669-677
• /
• 2020

As the number of Internet users exploded, attacks on the web increased. In addition, the attack patterns have been diversified to bypass existing defense techniques. Traditional web firewalls are difficult to detect attacks of unknown patterns.Therefore, the method of detecting abnormal behavior by artificial intelligence has been studied as an alternative. Specifically, attempts have been made to apply natural language processing techniques because the type of script or query being exploited consists of text. However, because there are many unknown words in scripts and queries, natural language processing requires a different approach. In this paper, we propose a new classification model which uses byte pair encoding (BPE) technology to learn the embedding vector, that is often used for web attack payloads, and uses an attention mechanism-based Bi-GRU neural network to extract a set of tokens that learn their order and importance. For major web attacks such as SQL injection, cross-site scripting, and command injection attacks, the accuracy of the proposed classification method is about 0.9990 and its accuracy outperforms the model suggested in the previous study.

• Journal of the Korea Society of Computer and Information
• /
• v.28 no.2
• /
• pp.201-207
• /
• 2023

In this paper, we propose an optimal model for mid to long-term price prediction of agricultural products using LGBM, MLP, LSTM, and GRU to compare and analyze the three strategies of the Multi-Step Time Series. The proposed model is designed to find the optimal combination between the models by selecting methods from various angles. Prior agricultural product price prediction studies have mainly adopted traditional econometric models such as ARIMA and LSTM-type models. In contrast, agricultural product price prediction studies related to Multi-Step Time Series were minimal. In this study, the experiment was conducted by dividing it into two periods according to the degree of volatility of agricultural product prices. As a result of the mid-to-long-term price prediction of three strategies, namely direct, hybrid, and multiple outputs, the hybrid approach showed relatively superior performance. This study academically and practically contributes to mid-to-long term daily price prediction by proposing an effective alternative.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.22 no.6
• /
• pp.165-171
• /
• 2022

Human heart rate can be used to measure exercise intensity as an important indicator. If heart rate can be predicted, exercise can be performed more efficiently by regulating the intensity of exercise in advance. In this paper, a FitRec-based prediction model is proposed for estimating running heart rate for users. Endomondo data is utilized for training the proposed prediction model. The processing algorithms for time-series data, such as LSTM(long short term memory) and GRU(gated recurrent unit), are employed to compare their performance. On the basis of simulation results, it was demonstrated that the proposed model trained with running exercise performed better than the model trained with several cardiac exercises.

• Journal of Bio-Environment Control
• /
• v.31 no.3
• /
• pp.152-162
• /
• 2022

This study developed simulation model for predicting the greenhouse interior environment using artificial intelligence machine learning techniques. Various methods have been studied to predict the internal environment of the greenhouse system. But the traditional simulation analysis method has a problem of low precision due to extraneous variables. In order to solve this problem, we developed a model for predicting the temperature inside the greenhouse using machine learning. Machine learning models are developed through data collection, characteristic analysis, and learning, and the accuracy of the model varies greatly depending on parameters and learning methods. Therefore, an optimal model derivation method according to data characteristics is required. As a result of the model development, the model accuracy increased as the parameters of the hidden unit increased. Optimal model was derived from the GRU algorithm and hidden unit 6 (r² = 0.9848 and RMSE = 0.5857℃). Through this study, it was confirmed that it is possible to develop a predictive model for the temperature inside the greenhouse using data outside the greenhouse. In addition, it was confirmed that application and comparative analysis were necessary for various greenhouse data. It is necessary that research for development environmental control system by improving the developed model to the forecasting stage.