• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.50 no.4
• /
• pp.259-267
• /
• 2022

The existing method of deriving the fail-safe design requirements for the domestic developed rotary-wing aircraft system may miss the factors that cause critical system function failures, when being applied to the latest integrated avionics system. It is because the existing method analyzes the severity effect of the failures caused by a single item. To solve the issue, we present a systematic analysis procedure for deriving fail-safe design requirements of system architecture by utilizing functional hazard assessment and development assurance level analysis of SAE ARP4754A, international standard for complex system development. To demonstrate that our proposed procedure can be a solution for the aforementioned issue, we set up experimental environments that include common factors that can cause critical function failures of a system, and we conducted a cross-validation with the existing method. As a result, we showed that the proposed procedure can identify the potential critical common factors that the existing method have missed, and that the proposed procedure can derive fail-safe design requirements to control the common factors.

• Phonetics and Speech Sciences
• /
• v.15 no.3
• /
• pp.75-82
• /
• 2023

Automatic speech recognition (ASR) has been revolutionized with deep learning-based approaches, among which self-supervised learning methods have proven to be particularly effective. In this study, we aim to enhance the performance of OpenAI's Whisper model, a multilingual ASR system on the Korean language. Whisper was pretrained on a large corpus (around 680,000 hours) of web speech data and has demonstrated strong recognition performance for major languages. However, it faces challenges in recognizing languages such as Korean, which is not major language while training. We address this issue by fine-tuning the Whisper model with an additional dataset comprising about 1,000 hours of Korean speech. We also compare its performance against a Transformer model that was trained from scratch using the same dataset. Our results indicate that fine-tuning the Whisper model significantly improved its Korean speech recognition capabilities in terms of character error rate (CER). Specifically, the performance improved with increasing model size. However, the Whisper model's performance on English deteriorated post fine-tuning, emphasizing the need for further research to develop robust multilingual models. Our study demonstrates the potential of utilizing a fine-tuned Whisper model for Korean ASR applications. Future work will focus on multilingual recognition and optimization for real-time inference.

• Journal of the Korean Society for information Management
• /
• v.40 no.4
• /
• pp.403-428
• /
• 2023

This study investigates the development path and intellectual structure of data literacy research, aiming to identify emerging topics in the field. A comprehensive search for data literacy-related articles on the Web of Science reveals that the field is primarily concentrated in Education & Educational Research and Information Science & Library Science, accounting for nearly 60% of the total. Citation network analysis, employing the PageRank algorithm, identifies key papers with high citation impact across various topics. To accurately trace the development path of data literacy research, an enhanced PageRank main path algorithm is developed, which overcomes the limitations of existing methods confined to the Education & Educational Research field. Keyword bibliographic coupling analysis is employed to unravel the intellectual structure of data literacy research. Utilizing the PNNC algorithm, the detailed structure and clusters of the derived keyword bibliographic coupling network are revealed, including two large clusters, one with two smaller clusters and the other with five smaller clusters. The growth index and mean publishing year of each keyword and cluster are measured to pinpoint emerging topics. The analysis highlights the emergence of critical data literacy for social justice in higher education amidst the ongoing pandemic and the rise of AI chatbots. The enhanced PageRank main path algorithm, developed in this study, demonstrates its effectiveness in identifying parallel research streams developing across different fields.

• Journal of the Microelectronics and Packaging Society
• /
• v.29 no.4
• /
• pp.21-27
• /
• 2022

Flexible pressure sensor was developed using low-cost conductive graphite as printed electronics. Flexible pressure sensors are attracting attention as materials to be used in future industries such as medical, games, and AI. As a result of evaluating various electromechanical properties of the printed electrode for flexible pressure sensors, it showed a constant resistance change rate in a maximum tensile rate of 20%, 30° tension/bending, and a simple pulse test. A more appropriate matrix pattern was designed by simulating the electrodes for which this verification was completed. Utilizing the Serpentine pattern, we utilized a process that allows simultaneous fabrication and encapsulation of the matrix pattern. One side of the printed graphite electrode was O₂ plasma surface treated to increase adhesive strength, rotated 90 times, and two electrodes were made into one through a lamination process. As a result of pasting the matrix pattern prepared in this way to the wrist pulse position of the human body and proceeding with the actual measurement, a constant rate of resistance change was shown regardless of gender.

• Korean Journal of Remote Sensing
• /
• v.39 no.6_3
• /
• pp.1693-1705
• /
• 2023

Rural areas, which account for about 90% of the country's land area, are increasing in importance and value as a space that performs various public functions. However, facilities that adversely affect residents' lives, such as livestock facilities, factories, and solar panels, are being built indiscriminately near residential areas, damaging the rural environment and landscape and lowering the quality of residents' lives. In order to prevent disorderly development in rural areas and manage rural space in a planned manner, detection and monitoring of hazardous facilities in rural areas is necessary. Data can be acquired through satellite imagery, which can be acquired periodically and provide information on the entire region. Effective detection is possible by utilizing image-based deep learning techniques using convolutional neural networks. Therefore, U-Net model, which shows high performance in semantic segmentation, was used to classify potentially hazardous facilities in rural areas. In this study, KOMPSAT ortho-mosaic optical imagery provided by the Korea Aerospace Research Institute in 2020 with a spatial resolution of 0.7 meters was used, and AI training data for livestock facilities, factories, and solar panels were produced by hand for training and inference. After training with U-Net, pixel accuracy of 0.9739 and mean Intersection over Union (mIoU) of 0.7025 were achieved. The results of this study can be used for monitoring hazardous facilities in rural areas and are expected to be used as basis for rural planning.

• KIPS Transactions on Software and Data Engineering
• /
• v.12 no.10
• /
• pp.445-454
• /
• 2023

Recently, various fault diagnosis studies are being conducted utilizing data from collaborative robots. Existing studies performing fault diagnosis on collaborative robots use static data collected based on the assumed operation of predefined devices. Therefore, the fault diagnosis model has a limitation of increasing dependency on the learned data patterns. Additionally, there is a limitation in that a diagnosis reflecting the characteristics of collaborative robots operating with multiple joints could not be conducted due to experiments using a single motor. This paper proposes an LSTM diagnostic model that can overcome these two limitations. The proposed method selects representative normal patterns using the correlation analysis of vibration and current data in single-axis and multi-axis work environments, and generates residual patterns through differences from the normal representative patterns. An LSTM model that can perform gear wear diagnosis for each axis is created using the generated residual patterns as inputs. This fault diagnosis model can not only reduce the dependence on the model's learning data patterns through representative patterns for each operation, but also diagnose faults occurring during multi-axis operation. Finally, reflecting both internal and external data characteristics, the fault diagnosis performance was improved, showing a high diagnostic performance of 98.57%.

• Journal of the Korea Society of Computer and Information
• /
• v.29 no.1
• /
• pp.273-285
• /
• 2024

With the active utilization of Online Judge (OJ) systems in the field of education, various studies utilizing learner data have emerged. This research proposes a problem recommendation based on a user-based collaborative filtering approach with learner data to support learners in their problem selection. Assistance in learners' problem selection within the OJ system is crucial for enhancing the effectiveness of education as it impacts the learning path. To achieve this, this system identifies learners with similar problem-solving tendencies and utilizes their problem-solving history. The proposed technique has been implemented on an OJ site in the fields of algorithms and programming, operated by the Chungbuk Education Research and Information Institute. The technique's service utility and usability were assessed through expert reviews using the Delphi technique. Additionally, it was piloted with site users, and an analysis of the ratio of correctness revealed approximately a 16% higher submission rate for recommended problems compared to the overall submissions. A survey targeting users who used the recommended problems yielded a 78% response rate, with the majority indicating that the feature was helpful. However, low selection rates of recommended problems and low response rates within the subset of users who used recommended problems highlight the need for future research focusing on improving accessibility, enhancing user feedback collection, and diversifying learner data analysis.

• Asia-Pacific Journal of Business Venturing and Entrepreneurship
• /
• v.19 no.3
• /
• pp.173-186
• /
• 2024

This paper analyzes a case of successful faculty entrepreneurship through a coupled process of open innovation in a university context, using the core competency theory perspective. Initially, the current state of faculty entrepreneurship is examined, and the effects of interdisciplinary coupled processes of open innovation are explored, focusing on the case of 'Omotion Inc.,' a startup utilizing generative AI technology for hyper-realistic 3D virtual human experiences. The research methodology involves in-depth interviews with Omotion Inc.'s co-founders, technology commercialization professionals, and experts in the field, followed by analysis based on foundational theories. Applying the core competency theory, this paper scrutinizes the process of integrating diverse expertise and technologies from various academic disciplines. The analysis goes beyond the limitations of faculty entrepreneurship confined to a single technology-centric research domain. Instead, it explores the possibilities of enhancement and value creation through coupled processes, providing practical implications for the university entrepreneurial ecosystem. The aim is to extend the traditional roles of education and research within the university, presenting a role in economic value creation beyond the boundaries of conventional faculty entrepreneurship. Through the collaboration of two faculty members, this study showcases the creation of novel technology and business models. It establishes that successful coupled processes of open innovation in faculty entrepreneurship, from a core competency theory perspective, require the entrepreneurial firm to possess (1) entrepreneurial capabilities, (2) technological capabilities, and (3) networking capabilities. The implications of this research highlight the positive impact of coupled processes of open innovation in faculty entrepreneurship, as evidenced by the Omotion Inc. case, offering guidance on entrepreneurial directions for university members preparing for entrepreneurship.

• Journal of Intelligence and Information Systems
• /
• v.29 no.3
• /
• pp.267-286
• /
• 2023

The conversational agents such as AI speakers utilize voice conversation for human-computer interaction. Voice recognition errors often occur in conversational situations. Recognition errors in user utterance records can be categorized into two types. The first type is misrecognition errors, where the agent fails to recognize the user's speech entirely. The second type is misinterpretation errors, where the user's speech is recognized and services are provided, but the interpretation differs from the user's intention. Among these, misinterpretation errors require separate error detection as they are recorded as successful service interactions. In this study, various text separation methods were applied to detect misinterpretation. For each of these text separation methods, the similarity of consecutive speech pairs using word embedding and document embedding techniques, which convert words and documents into vectors. This approach goes beyond simple word-based similarity calculation to explore a new method for detecting misinterpretation errors. The research method involved utilizing real user utterance records to train and develop a detection model by applying patterns of misinterpretation error causes. The results revealed that the most significant analysis result was obtained through initial consonant extraction for detecting misinterpretation errors caused by the use of unregistered neologisms. Through comparison with other separation methods, different error types could be observed. This study has two main implications. First, for misinterpretation errors that are difficult to detect due to lack of recognition, the study proposed diverse text separation methods and found a novel method that improved performance remarkably. Second, if this is applied to conversational agents or voice recognition services requiring neologism detection, patterns of errors occurring from the voice recognition stage can be specified. The study proposed and verified that even if not categorized as errors, services can be provided according to user-desired results.

• Food Science of Animal Resources
• /
• v.34 no.5
• /
• pp.582-590
• /
• 2014

The effect of mugwort extract (ME) and vitamin C (VC), added individually or in combination, on color, lipid oxidation, and sensory characteristics of chicken nuggets stored for 12 d was investigated. Eight treatments of chicken nuggets contained the following: Control (no antioxidant added), VC (0.05% VC), ME 0.05 (0.05% ME), ME 0.1 (0.1% ME), ME 0.2 (0.2% ME), VC+ME 0.05 (0.05% VC + 0.05% ME) and VC+ME 0.1 (0.05% VC + 0.1% ME), VC+ME 0.2 (0.05% VC + 0.2% ME). Results showed that the mixture of 0.05% VC and 0.2% ME was most effective for delaying lipid oxidation (thiobarbituric acid reactive substances, conjugated dienies, and peroxide formation) when compared to the control or ME alone added. The color values of all treatments were significantly affected by adding ME. Additionally, the total color difference (${\Delta}E$), chroma ($C^*$), and hue angle ($H^{\circ}$) values of all treatments, except for VC, were lower than those of the control as the amount of ME increased. The sensory characteristics (flavor, odor, and overall acceptability) did not differ significantly in any of the chicken nugget samples, whereas storage time had a significant effect. The results suggest that the possibility of utilizing chicken nuggets with a mixture of mugwort extract and vitamin C for the increase of shelf-life and quality.