• Journal of Intelligence and Information Systems
• /
• v.28 no.2
• /
• pp.191-206
• /
• 2022

Recently, it is a de-facto approach to utilize a pre-trained language model(PLM) to achieve the state-of-the-art performance for various natural language tasks(called downstream tasks) such as sentiment analysis and question answering. However, similar to any other machine learning method, PLM tends to depend on the data distribution seen during the training phase and shows worse performance on the unseen (Out-of-Distribution) domain. Due to the aforementioned reason, there have been many efforts to develop domain-specified PLM for various fields such as medical and legal industries. In this paper, we discuss the training of a finance domain-specified PLM for the Korean language and its applications. Our finance domain-specified PLM, KB-BERT, is trained on a carefully curated financial corpus that includes domain-specific documents such as financial reports. We provide extensive performance evaluation results on three natural language tasks, topic classification, sentiment analysis, and question answering. Compared to the state-of-the-art Korean PLM models such as KoELECTRA and KLUE-RoBERTa, KB-BERT shows comparable performance on general datasets based on common corpora like Wikipedia and news articles. Moreover, KB-BERT outperforms compared models on finance domain datasets that require finance-specific knowledge to solve given problems.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.21 no.6
• /
• pp.155-161
• /
• 2021

Gesture recognition analytics through a camera in real time have been widely studied in recent years. Since a small number of features from human joints are extracted, low accuracy of classifying models is get in conventional gesture recognition studies. In this paper, CBAM (Convolutional Block Attention Module) with high accuracy for classifying images is proposed as a classification model and algorithm calculating the angle of joints depending on actions is presented to solve the issues. Employing five exercise gestures images from the fitness posture images provided by AI Hub, the images are applied to the classification model. Important 8-joint angles information for classifying the exercise gestures is extracted from the images by using MediaPipe, a graph-based framework provided by Google. Setting the features as input of the classification model, the classification model is learned. From the simulation results, it is confirmed that the exercise gestures are classified with high accuracy in the proposed model.

• Journal of Industrial Convergence
• /
• v.22 no.8
• /
• pp.1-17
• /
• 2024

Recently, interest in edutech has been focused on its fusion with AI technology, and the market in this field is expanding. This study aims to analyze the technological competitiveness and key technological areas of major countries in the AI-edutech field. Additionally, considering that AI-edutech is a convergence of AI technology and edutech, the study seeks to examine the path dependence of AI-edutech in each country to determine whether they are based on existing AI technologies or edutech. To this end, AI-edutech patents were collected and competitiveness was analyzed using patent activity, patent impact, and market acquisition indicators. Path dependence for each country was analyzed using the hurdle negative binomial regression model. The analysis results indicate that the major countries in the AI-edutech field are China, South Korea, the United States, India, and Japan. In terms of patent activity, China had the highest level, followed by South Korea. In terms of patent impact and market securing power, the United States was high in both aspects, Japan had high market securing power, and South Korea had high patent influence. The results of the hurdle negative binomial analysis presented unique findings. The logit part results indicated that the possession of existing AI and edutech did not positively affect the emergence of current AI-edutech, but the count part results showed a positive influence. This suggests that, overall, it is difficult to assert that current AI-edutechs are based on past AI and edutechs. However, once some AI-edutechs based on existing AI and edutechs emerge, they are influenced by the existing technologies. These findings provide implications for future research and technological strategies in this field.

• Journal of Internet Computing and Services
• /
• v.23 no.1
• /
• pp.77-86
• /
• 2022

The Republic of Korea is a country in which coastal surveillance is an imperative national task as it is surrounded by seas on three sides under the confrontation between South and North Korea. However, due to Defense Reform 2.0, the number of R/D (Radar) operating personnel has decreased, and the period of service has also been shortened. Moreover, there is always a possibility that a human error will occur. This paper presents specific guidelines for developing an AI learning model for the intelligent coastal surveillance system. We present a three-step strategy to realize the guidelines. The first stage is a typical stage of building an AI learning model, including data collection, storage, filtering, purification, and data transformation. In the second stage, R/D signal analysis is first performed. Subsequently, AI learning model development for classifying real and false images, coastal area analysis, and vulnerable area/time analysis are performed. In the final stage, validation, visualization, and demonstration of the AI learning model are performed. Through this research, the first achievement of making the existing weapon system intelligent by applying the application of AI technology was achieved.

• Journal of Intelligence and Information Systems
• /
• v.29 no.1
• /
• pp.143-174
• /
• 2023

With the development of deep learning technologies, Artificial Intelligence powered Optical Character Recognition (AI-OCR) has evolved to read multiple languages from various forms of images accurately. For the financial industry, where a large number of diverse documents are processed through manpower, the potential for using AI-OCR is great. In this study, we present a configuration and a design of an AI-OCR modality for use in the financial industry and discuss the platform construction with application cases. Since the use of financial domain data is prohibited under the Personal Information Protection Act, we developed a deep learning-based data generation approach and used it to train the AI-OCR models. The AI-OCR models are trained for image preprocessing, text recognition, and language processing and are configured as a microservice architected platform to process a broad variety of documents. We have demonstrated the AI-OCR platform by applying it to financial domain tasks of document sorting, document verification, and typing assistance The demonstrations confirm the increasing work efficiency and conveniences.

• Annual Conference of KIPS
• /
• 2022.11a
• /
• pp.295-296
• /
• 2022

작물을 경작 해충과 질병은 오랫동안 주요 관심사였다. 농업에서 병해충을 탐지하기 위해 전통적인 방법을 사용하는 것은 더 이상 높은 효율성을 제공하지 않는다. 오늘날 과학과 인공 지능의 폭발적인 발달로 인해 농업분야의 연구원들은 병해충을 탐지하기 위해 딥 러닝을 적용하고 있다. 최근에 다양한 분야의 문제들을 해결하기 위해 수많은 모델들이 발표되었지만, 많은 병해충 진단 딥러닝을 사용한 방법들은 하드웨어 리소스를 낭비하고 실제 농장에서 사용하기 어렵다. 따라서 본 논문에서는 작물의 병해충을 분류하기 위해 Select Kernel Attention(SK Attention)을 Channel Attention Module 로 변경하여 Decoupling-and-Attention network (DANet)을 하드웨어 리소스 사용을 최소화한다.

• Journal of Advanced Navigation Technology
• /
• v.25 no.6
• /
• pp.576-581
• /
• 2021

Compared to a low concentration, a high concentration clearly entails limitations in terms of predictive performance owing to differences in its frequency and environment of occurrence. To resolve this problem, in this study, an artificial intelligence neural network algorithm was used to classify low and high concentrations; furthermore, two prediction models trained using the characteristics of the classified concentration types were used for prediction. To this end, we constructed training datasets using weather and air pollutant data collected over a decade in the Cheonan region. We designed a DNN-based classification model to classify low and high concentrations; further, we designed low- and high-concentration prediction models to reflect characteristics by concentration type based on the low and high concentrations classified through the classification model. According to the results of the performance assessment of the prediction model by concentration type, the low- and high-concentration prediction accuracies were 90.38% and 96.37%, respectively.

• Annual Conference on Human and Language Technology
• /
• 2018.10a
• /
• pp.47-49
• /
• 2018

최근 존재하는 대부분의 관계 추출 모델은 언급 수준의 관계 추출 모델이다. 이들은 성능은 높지만, 문서에 존재하는 다수의 문장을 처리할 때, 문서 내에 주요 개체 및 여러 문장에 걸쳐서 표현되는 개체간의 관계를 분류하지 못한다. 이는 높은 수준의 관계를 정의하지 못함으로써 올바르게 데이터를 정형화지 못하는 중대한 문제이다. 해당 논문에서는 이러한 문제를 타파하기 위하여 여러 문장에 걸쳐서 개체간의 상호작용 관계도 파악하는 전역 수준의 관계 추출 모델을 제안한다. 제안하는 모델은 전처리 단계에서 문서를 분석하여 사전 지식베이스, 개체 연결 그리고 각 개체의 언급횟수를 파악하고 문서 내의 주요 개체들을 파악한다. 이후 언급 수준의 관계 추출을 통하여 1차적으로 단편적인 관계 추출을 실행하고, 주요개체와 관련된 관계는 외부 메모리에 샘플로 저장한다. 이후 단편적 관계들과 외부메모리를 이용하여 여러 문장에 걸쳐 표현되는 개체 간 관계를 알아낸다. 해당 논문은 이러한 모델의 구조도와 실험방법의 설계에 대하여 설명하였고, 해당 실험의 기대효과 또한 작성하였다.

• Annual Conference on Human and Language Technology
• /
• 2018.10a
• /
• pp.661-664
• /
• 2018

온라인 게시판 글과 채팅창에서 주고받는 대화는 실제 사용되고 있는 구어체 특성이 잘 반영된 텍스트 코퍼스로 음성인식의 언어 모델 재료로 활용하기 좋은 학습 데이터이다. 하지만 온라인 특성상 노이즈가 많이 포함되어 있기 때문에 학습에 직접 활용하기가 어렵다. 본 논문에서는 사용자 입력오류가 다수 포함된 문장에서의 한글 오류 보정을 위한 sequence-to-sequence Denoising Autoencoder 모델을 제안한다.

• Journal of the Korea Society of Computer and Information
• /
• v.26 no.5
• /
• pp.39-46
• /
• 2021

AI is deeply applied to various algorithms that assists us, not only daily technologies like translator and Face ID, but also contributing to innumerable fields in industry, due to its dominance. In this research, we provide convenience through AI categorization, extracting the only data that users need, with objective classification, rather than verifying all data to find from the internet, where exists an immense number of contents. In this research, we propose a model using LSTM(Long-Short Term Memory Network), which stands out from text classification, and compare its performance with models of RNN(Recurrent Neural Network) and BiLSTM(Bidirectional LSTM), which is suitable structure for natural language processing. The performance of the three models is compared using measurements of accuracy, precision, and recall. As a result, the LSTM model appears to have the best performance. Therefore, in this research, text classification using LSTM is recommended.