• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.26 no.1
• /
• pp.70-75
• /
• 2022

Accurately measuring location is necessary to provide a variety of services. The data for indoor positioning measures the RSSI values from the WiFi device through an application of a smartphone. The measured data becomes the raw data of machine learning. The feature data is the measured RSSI value, and the label is the name of the space for the measured position. For this purpose, the machine learning technique is to study a technique that predicts the exact location only with the WiFi signal by applying an efficient technique to classification. Ensemble is a technique for obtaining more accurate predictions through various models than one model, including backing and boosting. Among them, Boosting is a technique for adjusting the weight of a model through a modeling result based on sampled data, and there are various algorithms. This study uses Xgboost among the above techniques and evaluates performance with other ensemble techniques.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.23 no.4
• /
• pp.115-119
• /
• 2023

The natural language processing models used in current artificial intelligence are huge, causing various difficulties in processing and analyzing data in real time. In order to solve these difficulties, we proposed a method to improve the efficiency of processing by using less memory and checked the performance of the proposed model. The technique applied in this paper to evaluate the performance of the proposed model is to divide the large corpus by adjusting the number of attention heads and embedding size of the BERT[1] model to be small, and the results are calculated by averaging the output values of each forward. In this process, a random offset was assigned to the sentences at every epoch to provide diversity in the input data. The model was then fine-tuned for classification. We found that the split processing model was about 12% less accurate than the unsplit model, but the number of parameters in the model was reduced by 56%.

• Environmental and Resource Economics Review
• /
• v.31 no.4
• /
• pp.481-503
• /
• 2022

This study examines how eco-innovation contributes to the exports of environmentally-friendly products using the dynamic panel model. The results reveal that the adjustment in the exports exists to recover the long-run equilibrium with sluggish adjustment speed. In addition, the results show that environmental patent applications and environment-related R&D expenditures are beneficial for enhancing the environmentally-friendly exports. While the environmental patent applications are associated only with an increase in the exports of products for resource management, the environmental R&D expenditures contribute to the exports of pollution management products, cleaner technologies and products, and resource management products. Moreover, as the long-run effects of eco-innovation on the exports become greater than the short-run effects, it appears that public eco-innovation is more likely to support future exports than private eco-innovation.

• Journal of the Korea Convergence Society
• /
• v.13 no.1
• /
• pp.189-199
• /
• 2022

While modern illustration designs require originality, the shape and color of marine life provide rich materials for illustration creation. Illustration has a very high artistic creation value. Based on the knowledge of color science and visual form design, this study takes illustration design and marine biology as the main research object. The purpose of this study is: ① The typical features of Marine life are sorted by color and form. ② Lead out graphic symbols representing Marine life. ③ Combine Marine life with illustration design to make design cases. Through research, it is highlighted that the curve feature is a typical feature for distinguishing marine organisms from terrestrial organisms. The classification and comparative analysis of the color phenomena of tropical fish help to extract and process the color of marine tropical fish on the basis of understanding the color characteristics of tropical fish, and apply it to the illustration design.

• The Journal of Bigdata
• /
• v.7 no.2
• /
• pp.11-29
• /
• 2022

Pre-trained language models are the most important and widely used tools in natural language processing tasks. Since those have been pre-trained for a large amount of corpus, high performance can be expected even with fine-tuning learning using a small number of data. Since the elements necessary for implementation, such as a pre-trained tokenizer and a deep learning model including pre-trained weights, are distributed together, the cost and period of natural language processing has been greatly reduced. Transformer variants are the most representative pre-trained language models that provide these advantages. Those are being actively used in other fields such as computer vision and audio applications. In order to make it easier for researchers to understand the pre-trained language model and apply it to natural language processing tasks, this paper describes the definition of the language model and the pre-learning language model, and discusses the development process of the pre-trained language model and especially representative Transformer variants.

• Smart Media Journal
• /
• v.12 no.6
• /
• pp.15-26
• /
• 2023

As face-to-face communication has become difficult due to the COVID-19 pandemic, studies have been conducted to understand the impact of non-face-to-face communication, but there is a lack of research that examines this through messenger application reviews. This study aims to identify the impact of the pandemic through Latent Dirichlet Allocation (LDA) topic modeling by collecting review data of 메신저 applications in the Google Play Store and suggest service strategies accordingly. The study categorized the data based on when the pandemic started and the ratings given by users. The analysis showed that messenger is mainly used by middle-aged and older people, and that family communication increased after the pandemic. Users expressed frustration with the application's updates and found it difficult to adapt to the changes. This calls for a development approach that adjusts the frequency of updates and actively listens to user feedback. Also, providing an intuitive and simple user interface (UI) is expected to improve user satisfaction.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2022.05a
• /
• pp.233-233
• /
• 2022

최근 기후변화로 인해 집중호우 및 돌발홍수의 증가로 침수피해가 빈번하게 발생하고 있다. 마찬가지로 해안지역의 피해 또한 증가하고 있으나, 해안지역의 특성을 고려한 연구가 미비한 실정이다. 따라서 본 연구에서 해안지역의 특성을 고려해 폭풍해일로 인한 월파뿐만 아니라 강우도 고려하여 해안지역의 범람 양상을 확인하고자 하였다. 본 연구에서는 국내 해안지역에 대한 빈도별 폭풍해일과 강우로인한 범람 모의를 진행하였다. 우선, 수치해석 모형의 경계조건을 산정하기 위해 EurOtop(2018)의 경험식을 이용하여 월파량을 산정하였다. EurOtop의 월파량 산정 시 암석 옹벽이 아닌 콘크리트 옹벽으로된 경사식 단면으로 고려하여 계산하였고 산책로와 벽까지 고려하여 계산하였다. 경험식 계산을 위해 매개변수(유의파고, 여유고, 구조물의 조도계수, 구조물의 기울기 및 경사 등)를 조정하여 계산하였다. 이 중, 계산에 사용된 유의파고는 시나리오별 강우에 대해 SWAN(Simulating WAves Nearshore)으로 계산된 값을 활용하였고, 해안선을 두 부분으로 나누어 해안지역 각 지점별 파고값의 평균을 사용해 월파량 계산을 진행했다. 이때, 파고의 종류로 5% 확률의 파고, 평균 파고, 중앙값 파고, 95% 확률의 파고로 분류해 월파량 계산을 진행했고, 그 중, 평균 파고를 이용해 계산한 월파량을 수치해석 모델의 입력자료로 활용하였다. 시나리오별로 계산된 월파량만을 이용해 2차원 침수모형인 FLO-2D의 경계조건 입력값으로 사용하여 침수 양상을 표출하기 위해 Mapper와 ArcGIS를 이용하여 침수와 범람 양상을 확인하였다. 또, 다른 조건으로 시나리오별 계산된 월파량, 연구유역 해안 반대편에 위치한 산으로부터 유입되는 물의 양 그리고 해안지역 전체에 내리는 강우를 입력자료로 사용해 모의를 진행한 후 Mapper와 ArcGIS로 표출하여 침수 및 범람 양상을 확인하였다.

• Tunnel and Underground Space
• /
• v.33 no.5
• /
• pp.299-311
• /
• 2023

Continuous excavation technologies are developed to improve the excavation rate of shield TBM. Continuous excavation is a technology that provides thrust to segments, excluding being installed one, to reduce tunneling downtime. This paper investigated the prior technology related to continuous excavation segments. The main technology was classified into helical segment, honeycomb segment, and conventional segment methods. The helical segment method has not been applied in actual construction yet, and the honeycomb segment method has not succeeded in commercialization. The continuous excavation method using conventional segments has been successfully demonstrated. The thrust force and operation method of the thrust jacks for the semi-continuous technology were analyzed. Continuous excavation TBM research is also progressing in Korea, and through the analysis of successful cases, the need to develop independent continuous excavation methods has been identified.

• Journal of Korea Society of Industrial Information Systems
• /
• v.29 no.3
• /
• pp.79-87
• /
• 2024

Recently, the importance of pre-trained language models (PLM) has been emphasized for natural language processing (NLP) such as text classification, sentiment analysis, and question answering. Korean PLM shows high performance in NLP in general-purpose domains, but is weak in domains such as finance, medicine, and law. The main goal of this study is to propose a language model learning process and method to build a financial-specific language model that shows good performance not only in the financial domain but also in general-purpose domains. The five steps of the financial-specific language model are (1) financial data collection and preprocessing, (2) selection of model architecture such as PLM or foundation model, (3) domain data learning and instruction tuning, (4) model verification and evaluation, and (5) model deployment and utilization. Through this, a method for constructing pre-learning data that takes advantage of the characteristics of the financial domain and an efficient LLM training method, adaptive learning and instruction tuning techniques, were presented.

• Journal of Korea Society of Industrial Information Systems
• /
• v.29 no.1
• /
• pp.1-13
• /
• 2024

This study aims to provide a customized dynamic kiosk screen that considers user characteristics to cope with changes caused by increased use of kiosks. In order to optimize the screen composition according to the characteristics of the digital vulnerable group such as the visually impaired, the elderly, children, and wheelchair users, etc., users are classified into nine categories based on real-time analysis of user characteristics (wheelchair use, visual impairment, age, etc.). The kiosk screen is dynamically adjusted according to the characteristics of the user to provide efficient services. This study shows that the system communication and operation were performed in the embedded environment, and the used object detection, gait recognition, and speech recognition technologies showed accuracy of 74%, 98.9%, and 96%, respectively. The proposed technology was verified for its effectiveness by implementing a prototype, and through this, this study showed the possibility of reducing the digital gap and providing user-friendly "barrier-free kiosk" services.