• KIPS Transactions on Software and Data Engineering
• /
• v.12 no.10
• /
• pp.455-460
• /
• 2023

Text augmentation is a methodology that creates new augmented texts by transforming or generating original texts for the purpose of improving the performance of NLP models. However existing text augmentation techniques have limitations such as lack of expressive diversity semantic distortion and limited number of augmented texts. Recently text augmentation using large language models and few-shot learning can overcome these limitations but there is also a risk of noise generation due to incorrect generation. In this paper, we propose a text augmentation method called TAGS that generates multiple candidate texts and selects the appropriate text as the augmented text. TAGS generates various expressions using few-shot learning while effectively selecting suitable data even with a small amount of original text by using contrastive learning and similarity comparison. We applied this method to task-oriented chatbot data and achieved more than sixty times quantitative improvement. We also analyzed the generated texts to confirm that they produced semantically and expressively diverse texts compared to the original texts. Moreover, we trained and evaluated a classification model using the augmented texts and showed that it improved the performance by more than 0.1915, confirming that it helps to improve the actual model performance.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.23 no.5
• /
• pp.21-27
• /
• 2023

As a means of helping ships navigate safely, navigational aids in operation in the maritime envirionment require periodic management, and due to the nature of the environment, it is difficult to visually check the exact state. As a result, the smart navigation aid system, which improves route safety and operational efficiency, utillizes expertise including sensors, communications, and information technology, unlike general route markings. The communication environment of the smart navigation aid system, which aims to ensure the safety of the navigators operating the ship and the safety of the ship, uses a wireless communication network in accordance with the marine environment. The ship collects the information necessary for the maritime environment on the land and operates. In this process, there is a need to consider the wireless communication security guideline. Basically, based on IHO S-100 a standard for facilitating data exchange and SECOM, which provides an interface for safe communication. This paper research a security system for safe communication in a maritime environment. The security system for the basic interface based on the document was presented, and there were some vulnerabillties to data exchange due to the wireless communication characteristics of the maritime environment, and the user authetication part was added considering the vulnerability that unauthorized users can access the service.

• The Journal of the Convergence on Culture Technology
• /
• v.9 no.6
• /
• pp.169-179
• /
• 2023

The purpose of this study is to suggest the direction of secondary school career education in a lifelong learning society. In that direction, first, it is necessary to strengthen teacher capacity and develop professionalism. Second, career education programs need to be improved and diversified. Third, it is necessary to strengthen collaboration and communication with career education experts as a way to strengthen cooperation and connection outside the school. Fourth, it is necessary to support online career education through improvement of career information network. Fifth, there are policy support and institutional improvement plans. Sixth, it is necessary to expand the subject of career education to the entire life. To this end, career education in secondary school is designed to flexibly cope with changes, overcome crises and turn them into opportunities, and provide experiences to solve problems. Comprehensive support for individually customized career education from a lifelong perspective that manages the degree is needed. Second, it is necessary to expand field-oriented career and job experience to provide sufficient job-related experience and support mentoring by field experts. Third, it is necessary to establish a career education network where schools, education offices, and local communities work together. Fourth, retraining of career counseling teachers is also necessary to support the strengthening of career education capacity of all teachers.

• Journal of the Institute of Convergence Signal Processing
• /
• v.24 no.2
• /
• pp.103-111
• /
• 2023

Consensus, that aims to converge the states of agents to the same states through information exchanges between agents, has been widely studied to control the multi-agent systems. In real systems, the measurement variables of each agent may be different, the loss of information across communication may occur, and the different networks for each state may need to be constructed for safety. Moreover, the input saturation and the disturbances in the system may cause instability. Therefore, this paper studies the PID(Proportional-Integral-Derivative)-based consensus control to achieve the swarm behavior of the multi-agent systems considering the heterogeneous state information, the input saturations, and the disturbances. Specifically, we consider the multiple follower agents and the single leader agent modeled by the second-order systems, and investigate the conditions to achieve the consensus based on the stability of the error system. It is confirmed that the proposed algorithm can achieve the consensus if only the connectivity of the position graph is guaranteed. Moreover, by extending the consensus algorithm, we study the formation control problem for the multi-agent systems. Finally, the validity of the proposed algorithm was verified through the simulations.

• Journal of the Institute of Convergence Signal Processing
• /
• v.24 no.2
• /
• pp.119-125
• /
• 2023

In this study, we photograph driving obstacle objects such as crosswalks, side spheres, manholes, braille blocks, partial ramps, temporary safety barriers, stairs, and inclined curb that hinder or cause inconvenience to the movement of the vulnerable using electric mobility aids. We develop an optimal AI model that classifies photographed objects and automatically recognizes them, and implement an algorithm that can efficiently determine obstacles in front of electric mobility aids. In order to enable object detection to be AI learning with high probability, the labeling form is labeled as a polygon form when building a dataset. It was developed using a Mask R-CNN model in Detectron2 framework that can detect objects labeled in the form of polygons. Image acquisition was conducted by dividing it into two groups: the general public and the transportation weak, and image information obtained in two areas of the test bed was secured. As for the parameter setting of the Mask R-CNN learning result, it was confirmed that the model learned with IMAGES_PER_BATCH: 2, BASE_LEARNING_RATE 0.001, MAX_ITERATION: 10,000 showed the highest performance at 68.532, so that the user can quickly and accurately recognize driving risks and obstacles.

• Journal of the Institute of Convergence Signal Processing
• /
• v.24 no.1
• /
• pp.42-50
• /
• 2023

The development of technology has achieved digital innovation throughout the media industry, including production techniques and editing technologies, and has brought diversity in the form of consumer viewing through the OTT service and streaming era. The convergence of big data and deep learning networks automatically generated text in format such as news articles, novels, and scripts, but there were insufficient studies that reflected the author's intention and generated story with contextually smooth. In this paper, we describe the flow of pictures in the storyboard with image caption generation techniques, and the automatic generation of story-tailored scenarios through language models. Image caption using CNN and Attention Mechanism, we generate sentences describing pictures on the storyboard, and input the generated sentences into the artificial intelligence natural language processing model KoGPT-2 in order to automatically generate scenarios that meet the planning intention. Through this paper, the author's intention and story customized scenarios are created in large quantities to alleviate the pain of content creation, and artificial intelligence participates in the overall process of digital content production to activate media intelligence.

• Tunnel and Underground Space
• /
• v.33 no.6
• /
• pp.445-471
• /
• 2023

Long-term safety over millennia is the top priority consideration in the construction of disposal sites. However, ensuring the mechanical stability of deep geological repositories for spent fuel, a.k.a. radwaste, disposal during construction and operation is also crucial for safe operation of the repository. Imposing restrictions or limitations on tunnel support and lining materials such as shotcrete, concrete, grouting, which might compromise the sealing performance of backfill and buffer materials which are essential elements for the long-term safety of disposal sites, presents a highly challenging task for rock engineers and tunnelling experts. In this study, as part of an extensive exploration to aid in the proper selection of disposal sites, the anticipation of constructing a deep geological repository at a depth of 500 meters in an unknown state has been carried out. Through a review of 2D and 3D numerical analyses, the study aimed to explore the range of properties that ensure stability. Preliminary findings identified the potential range of rock properties that secure the stability of central and disposal tunnels, while the stability of the vertical tunnel network was confirmed through 3D analysis, outlining fundamental rock conditions necessary for the construction of disposal sites.

• Journal of the Economic Geographical Society of Korea
• /
• v.26 no.3
• /
• pp.223-237
• /
• 2023

The Seoul Onsu Industrial Complex, having been completed over 50 years ago, is an old industrial complex, with deteriorating infrastructure and factory buildings. Despite this, there's a current urban industrial agglomeration centered on the machinery industry in the Seoul Onsu Industrial Complex. This study aims to holistically analyze the physical deterioration of facilities in the aging industrial complex and the characteristics of industrial agglomeration to derive the identity of the Seoul Onsu Industrial Complex. Based on the research findings, the complex is seeing an enhanced urban industrial agglomeration due to the influx of small-scale businesses resulting from concentrated trade networks in the metropolitan area and plot subdivision, permission for noise-producing processes, and the ease of securing highly-skilled technicians. However, this agglomeration coexists with a weakening of the complex's production function, limited innovativeness of resident companies, and non-innovative features resulting from weakened competitiveness in the metropolitan machinery industry. In summary, the identity of the Seoul Onsu Industrial Complex is a 'Non-Innovative Urban Industry Agglomeration', an old industrial complex, witnessing non-innovative agglomeration based on a machinery industry network centered in the metropolitan area.

• Journal of Intelligence and Information Systems
• /
• v.29 no.4
• /
• pp.91-110
• /
• 2023

The importance of manuals is emphasized because chemical accidents require swift response and recovery, and often result in environmental pollution and casualties. In this regard, the OECD revised OECD Guidelines for the Prevention, Preparedness, and Response to Chemical Accidents (referred to as the OECD Guidelines), in June 2023. Moreover, while existing research primarily raises awareness about chemical accidents, highlighting the need for a system-wide response including laws, regulations, and manuals, it was difficult to find comparative research on the attributes of manuals. So, this paper aims to compare and analyze the second and third editions of the OECD Guidelines, in order to uncover the information attributes and implications of the revised version. Specifically, TF-IDF (Term Frequency-Inverse Document Frequency) was applied to understand which keywords have become more important, and Word2Vec was applied to identify keywords that were used similarly and those that were differentiated. Lastly, a 2×2 matrix was proposed, identifying the topics within each quadrant to provide a deeper comparison of the information attributes of the OECD Guidelines. This study offers a framework to help researchers understand information attributes. From a practical perspective, it appears valuable for the revision of standard manuals by domestic government agencies and corporations related to chemistry.

• Journal of agricultural medicine and community health
• /
• v.48 no.4
• /
• pp.262-274
• /
• 2023

Objectives: This study examined the evaluation and potential improvements of 'Integrated Healthcare and Social Welfare service model' based on the experiences of practitioners from institutions participating in the 'Dalgubeol Health Doctor Services' and the service recipients. Methods: Qualitative research was conducted from September to November 2022 in this study, focusing on 4 providers from the dedicated Dalgubeol Health Doctor Services Team, 5 contact partners from affiliated organizations, and 6 service beneficiaries. The data gathered underwent thematic analysis. Results: The evaluation indicated that Dalgubeol Health Doctor Services has proven to be effective in addressing the complex needs of vulnerable populations. By providing integrated services through quick and simple beneficiary selection and resource linkage, it has contributed to the resolution of complex demands, recovery of positive attitudes towards life, and improvement in quality of life for users who have fear the use of medical and welfare services. Dalgubeol Health Doctor Services has established an integrated health care system involving not only public but also private organizations, from the referral agency to the service provider. Centered around Daegu Medical Center and involving five tertiary hospitals, it has established a model that supports treatment appropriate to the severity of the patient, from mild to severe. Conclusions: These findings indicate an enhancement in health equity, achieved through the active identification and subsequent health and welfare issue resolution of individuals marginalized from medical benefits.