• Journal of Convergence for Information Technology
• /
• v.12 no.2
• /
• pp.47-53
• /
• 2022

Recently, with the change of the intelligent security paradigm, study to apply various information generated from various information security systems to AI-based anomaly detection is increasing. Therefore, in this study, in order to convert log-like time series data into a vector, which is a numerical feature, the CBOW and Skip-gram inference methods of deep learning-based Word2Vec model and statistical method based on the coincidence frequency were used to transform the published ADFA system call data. In relation to this, an experiment was carried out through conversion into various embedding vectors considering the dimension of vector, the length of sequence, and the window size. In addition, the performance of the embedding methods used as well as the detection performance were compared and evaluated through GRU-based anomaly detection model using vectors generated by the embedding model as an input. Compared to the statistical model, it was confirmed that the Skip-gram maintains more stable performance without biasing a specific window size or sequence length, and is more effective in making each event of sequence data into an embedding vector.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2022.10a
• /
• pp.87-89
• /
• 2022

Recently, as more consumers spend more time at home due to COVID-19, the time spent on digital consumption such as SNS and OTT, which can be easily used non-face-to-face, naturally increased. Since 2019, when COVID-19 occurred, digital consumption has doubled from 44% to 82%, and it is important to quickly and accurately grasp and apply trends by analyzing consumers' emotions due to the rapidly changing digital characteristics. However, there are limitations in actually implementing services using emotional analysis in small systems rather than large-scale systems, and there are not many cases where they are actually serviced. However, if even a small system can easily analyze consumer trends, it will help the rapidly changing modern society. In this paper, we propose a lightweight trend analysis system that builds a learning network through Transfer Learning (Fine Tuning) of the BERT Model and interlocks Crawler for real-time data collection.

• Journal of Korea Society of Industrial Information Systems
• /
• v.29 no.4
• /
• pp.77-89
• /
• 2024

Knee osteoarthritis (OA) is a very common musculoskeletal disorder worldwide. The assessment of knee osteoarthritis, which requires a rapid and accurate initial diagnosis, is determined to be different depending on the currently dispersed classification system, and each classification system has different criteria. Also, because the medical staff directly sees and reads the X-ray pictures, it depends on the subjective opinion of the medical staff, and it takes time to establish an accurate diagnosis and a clear treatment plan. Therefore, in this study, we designed the stenosis length measurement algorithm and Osteophyte detection and length measurement algorithm, which are the criteria for determining the knee osteoarthritis grade, separately using CNN, which is a deep learning technique. In addition, we would like to create a grading system that integrates and complements the existing classification system and show results that match the judgments of actual medical staff. Based on publicly available OAI (Osteoarthritis Initiative) data, a total of 9,786 knee osteoarthritis data were used in this study, eventually achieving an Accuracy of 69.8% and an F1 score of 76.65%.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.24 no.1
• /
• pp.65-74
• /
• 2024

In recent years, image processing technology has played a critical role in the field of autonomous driving. Among them, image recognition technology is essential for the safety and performance of autonomous vehicles. Therefore, this paper aims to develop a hybrid image recognition system to enhance the safety and performance of autonomous vehicles. In this paper, various image recognition technologies are utilized to construct a system that recognizes and tracks objects in the vehicle's surroundings. Machine learning and deep learning algorithms are employed for this purpose, and objects are identified and classified in real-time through image processing and analysis. Furthermore, this study aims to fuse image processing technology with vehicle control systems to improve the safety and performance of autonomous vehicles. To achieve this, the identified object's information is transmitted to the vehicle control system to enable appropriate autonomous driving responses. The developed hybrid image recognition system in this paper is expected to significantly improve the safety and performance of autonomous vehicles. This is expected to accelerate the commercialization of autonomous vehicles.

• Journal of Intelligence and Information Systems
• /
• v.22 no.2
• /
• pp.127-142
• /
• 2016

Deep learning model is a kind of neural networks that allows multiple hidden layers. There are various deep learning architectures such as convolutional neural networks, deep belief networks and recurrent neural networks. Those have been applied to fields like computer vision, automatic speech recognition, natural language processing, audio recognition and bioinformatics where they have been shown to produce state-of-the-art results on various tasks. Among those architectures, convolutional neural networks and recurrent neural networks are classified as the supervised learning model. And in recent years, those supervised learning models have gained more popularity than unsupervised learning models such as deep belief networks, because supervised learning models have shown fashionable applications in such fields mentioned above. Deep learning models can be trained with backpropagation algorithm. Backpropagation is an abbreviation for "backward propagation of errors" and a common method of training artificial neural networks used in conjunction with an optimization method such as gradient descent. The method calculates the gradient of an error function with respect to all the weights in the network. The gradient is fed to the optimization method which in turn uses it to update the weights, in an attempt to minimize the error function. Convolutional neural networks use a special architecture which is particularly well-adapted to classify images. Using this architecture makes convolutional networks fast to train. This, in turn, helps us train deep, muti-layer networks, which are very good at classifying images. These days, deep convolutional networks are used in most neural networks for image recognition. Convolutional neural networks use three basic ideas: local receptive fields, shared weights, and pooling. By local receptive fields, we mean that each neuron in the first(or any) hidden layer will be connected to a small region of the input(or previous layer's) neurons. Shared weights mean that we're going to use the same weights and bias for each of the local receptive field. This means that all the neurons in the hidden layer detect exactly the same feature, just at different locations in the input image. In addition to the convolutional layers just described, convolutional neural networks also contain pooling layers. Pooling layers are usually used immediately after convolutional layers. What the pooling layers do is to simplify the information in the output from the convolutional layer. Recent convolutional network architectures have 10 to 20 hidden layers and billions of connections between units. Training deep learning networks has taken weeks several years ago, but thanks to progress in GPU and algorithm enhancement, training time has reduced to several hours. Neural networks with time-varying behavior are known as recurrent neural networks or RNNs. A recurrent neural network is a class of artificial neural network where connections between units form a directed cycle. This creates an internal state of the network which allows it to exhibit dynamic temporal behavior. Unlike feedforward neural networks, RNNs can use their internal memory to process arbitrary sequences of inputs. Early RNN models turned out to be very difficult to train, harder even than deep feedforward networks. The reason is the unstable gradient problem such as vanishing gradient and exploding gradient. The gradient can get smaller and smaller as it is propagated back through layers. This makes learning in early layers extremely slow. The problem actually gets worse in RNNs, since gradients aren't just propagated backward through layers, they're propagated backward through time. If the network runs for a long time, that can make the gradient extremely unstable and hard to learn from. It has been possible to incorporate an idea known as long short-term memory units (LSTMs) into RNNs. LSTMs make it much easier to get good results when training RNNs, and many recent papers make use of LSTMs or related ideas.

• Journal of Ocean Engineering and Technology
• /
• v.33 no.6
• /
• pp.590-596
• /
• 2019

Numerous deaths and substantial property damage have occurred recently due to frequent disasters of the highest intensity according to the abnormal climate, which is caused by various problems, such as global warming, all over the world. Such large-scale disasters have become an international issue and have made people aware of the disasters so they can implement disaster-prevention measures. Extensive information on disaster prevention actively has been announced publicly to support the natural disaster reduction measures throughout the world. In Japan, diverse developmental studies on disaster prevention systems, which support hazard map development and flood control activity, have been conducted vigorously to estimate external forces according to design frequencies as well as expected maximum frequencies from a variety of areas, such as rivers, coasts, and ports based on broad disaster prevention data obtained from several huge disasters. However, the current reduction measures alone are not sufficiently effective due to the change of the paradigms of the current disasters. Therefore, in order to obtain the synergy effect of reduction measures, a study of the establishment of an integrated system is required to improve the various disaster prevention technologies and the current disaster prevention system. In order to develop a similar typhoon search system and establish a disaster prevention infrastructure, in this study, techniques will be developed that can be used to forecast typhoons before they strike by using artificial intelligence (AI) technology and offer primary disaster prevention information according to the direction of the typhoon. The main function of this model is to predict the most similar typhoon among the existing typhoons by utilizing the major typhoon information, such as course, central pressure, and speed, before the typhoon directly impacts South Korea. This model is equipped with a combination of AI and DNN forecasts of typhoons that change from moment to moment in order to efficiently forecast a current typhoon based on similar typhoons in the past. Thus, the result of a similar typhoon search showed that the quality of prediction was higher with the grid size of one degree rather than two degrees in latitude and longitude.

• Journal of Korean Library and Information Science Society
• /
• v.52 no.1
• /
• pp.155-178
• /
• 2021

In the era of the 4th industrial revolution, public libraries need a strategy for promoting intelligent library services in order to actively respond to changes in the external environment such as artificial intelligence. Therefore, in this study, based on the concept of artificial intelligence and analysis of domestic and foreign artificial intelligence related trends, policies, and cases, we proposed the future direction of introduction and development of artificial intelligence services in the library. Currently, the library operates a reference information service that automatically provides answers through the introduction of artificial intelligence technologies such as deep learning and natural language processing, and develops a big data-based AI book recommendation and automatic book inspection system to increase business utilization and provide customized services for users. Has been provided. In the field of companies and industries, regardless of domestic and overseas, we are developing and servicing technologies based on autonomous driving using artificial intelligence, personal customization, etc., and providing optimal results by self-learning information using deep learning. It is developed in the form of an equation. Accordingly, in the future, libraries will utilize artificial intelligence to recommend personalized books based on the user's usage records, recommend reading and culture programs, and introduce real-time delivery services through transport methods such as autonomous drones and cars in the case of book delivery service. Service development should be promoted.

• KIPS Transactions on Computer and Communication Systems
• /
• v.12 no.3
• /
• pp.103-110
• /
• 2023

With the spread of the Internet, anonymous communities emerged along with the activation of communities for communication between people, and many users are doing harm to others, such as posting aggressive posts and leaving comments using anonymity. In the past, administrators directly checked posts and comments, then deleted and blocked them, but as the number of community users increased, they reached a level that managers could not continue to monitor. Initially, word filtering techniques were used to prevent malicious writing from being posted in a form that could not post or comment if a specific word was included, but they avoided filtering in a bypassed form, such as using similar words. As a way to solve this problem, deep learning was used to monitor posts posted by users in real-time, but recently, the community uses words that can only be understood by the community or from a human perspective, not from a general Korean word. There are various types and forms of characters, making it difficult to learn everything in the artificial intelligence model. Therefore, in this paper, we proposes a preprocessing technique in which each character of a sentence is imaged using a CNN model that learns the consonants, vowel and spacing images of Korean word and converts characters that can only be understood from a human perspective into characters predicted by the CNN model. As a result of the experiment, it was confirmed that the performance of the LSTM, BiLSTM and CNN-BiLSTM models increased by 3.2%, 3.3%, and 4.88%, respectively, through the proposed preprocessing technique.

• Journal of KIISE
• /
• v.44 no.5
• /
• pp.503-509
• /
• 2017

In this paper, the copy mechanism and input feeding are applied to recurrent neural network(RNN)-search model in a Korean-document summarization in an end-to-end manner. In addition, the performances of the document summarizations are compared according to the model and the tokenization format; accordingly, the syllable-unit, morpheme-unit, and hybrid-unit tokenization formats are compared. For the experiments, Internet newspaper articles were collected to construct a Korean-document summary data set (train set: 30291 documents; development set: 3786 documents; test set: 3705 documents). When the format was tokenized as the morpheme-unit, the models with the input feeding and the copy mechanism showed the highest performances of ROUGE-1 35.92, ROUGE-2 15.37, and ROUGE-L 29.45.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2019.11a
• /
• pp.71-72
• /
• 2019

최근 공공안전 분야에서는 국민의 위험상황을 분석하여 선제적으로 예측을 하여 국민의 안전을 보장하기 위한 요구사항이 대두대고 있다. 또한 스마트폰 및 스마트워치와 같은 고성능 모바일 단말 기기들의 대중화로 인해 해당 기기들에 부착된 다양한 센서 데이터들을 융복합하여 분석할 경우, 수집한 센서 데이터의 잠재적 가치를 안전보장 측면에서 사용할 수 있는 장점이 있다. 본 논문에서는 대인, 대물, 장소에 해당하는 로그 데이터들을 융복합 분석하여 보호대상자의 안전을 지원하는 다중로그 플랫폼 기반 이동경로 분석 기법을 제안한다. 다중로그 플랫폼에서 수집하는 보호대상자의 이동 경로 궤적을 활용하여 과거에 축적된 이동경로 패턴과 비교를 통해 현재 경로가 평소에 이용하던 경로와의 유사도를 추천하게 된다. 해당 이동 경로 분석 시스템은 위치기반 멀티모달 센서 데이터를 융복합 하여 보호대상자의 안전을 보장하는데 기여 할 것으로 예상된다.