• Journal of the Korean Society of Safety
• /
• v.34 no.2
• /
• pp.48-55
• /
• 2019

For preventing the accidents generated from the chemical materials, thus far, MSDS (Material Safety Data Sheet) data have been made to notify how to use and manage the hazardous and chemical materials in safety. However, it is difficult for users who handle these materials to understand the MSDS data because they are only listed based on the alphabetical order, not based on the specific factors such as similarity of characteristics. It is limited in representing the types of chemical materials with respect to their characteristics. Thus, in this study, a lots of MSDS data are visualized based on relationships of the characteristics among the chemical materials for supporting safety managers. For this, we used the textmining algorithm which extracts text keywords contained in documents and the Self-Organizing Map (SOM) algorithm which visually addresses textual data information. In the case of Occupational Safety and Health Administration (OSHA) in the United States, the guide texts contained in MSDS documents, which include use information such as reactivity and potential risks of materials, are gathered as the target data. First, using the textmining algorithm, the information of chemicals is extracted from these guide texts. Next, the MSDS map is developed using SOM in terms of similarity of text information of chemical materials. The MSDS map is helpful for effectively classifying chemical materials by mapping prohibited and hazardous substances on the developed the SOM map. As a result, using the MSDS map, it is easy for safety managers to detect prohibited and hazardous substances with respect to the Industrial Safety and Health Act standards.

• Journal of The Korean Society of Agricultural Engineers
• /
• v.61 no.3
• /
• pp.55-65
• /
• 2019

Although crack on concrete exists from its early formation, crack requires attention as it affects stiffness of structure and can lead demolition of structure as it grows. Detecting cracks on concrete is needed to take action prior to performance degradation of structure, and deep learning can be utilized for it. In this study, transfer learning, one of the deep learning techniques, was used to detect the crack, as the amount of crack's image data was limited. Pre-trained Inception-v3 was applied as a base model for the transfer learning. Web scrapping was utilized to fetch images of concrete wall with or without crack from web. In the recognition of crack, image post-process including changing size or removing color were applied. In the visualization of crack, source images divided into 30px, 50px or 100px size were used as input data, and different numbers of input data per category were applied for each case. With the results of visualized crack image, false positive and false negative errors were examined. Highest accuracy for the recognizing crack was achieved when the source images were adjusted into 224px size under gray-scale. In visualization, the result using 50 data per category under 100px interval size showed the smallest error. With regard to the false positive error, the best result was obtained using 400 data per category, and regarding to the false negative error, the case using 50 data per category showed the best result.

• Journal of the Korean Society of Safety
• /
• v.36 no.5
• /
• pp.61-70
• /
• 2021

This study aimed to analyze the research trends of the abstract data of ergonomic studies registered in MEDLINE, a medical bibliographic database, using word embedding. Medical-related ergonomic studies mainly focus on work-related musculoskeletal disorders, and there are no studies on the analysis of words as data using natural language processing techniques, such as word embedding. In this study, the abstract data of ergonomic studies were extracted with a program written with selenium and BeutifulSoup modules using python. The word embedding of the abstract data was performed using the word2vec model, after which the data found in the abstract were vectorized. The vectorized data were visualized in two dimensions using t-Distributed Stochastic Neighbor Embedding (t-SNE). The word "ergonomics" and ten of the most frequently used words in the abstract were selected as keywords. The results revealed that the most frequently used words in the abstract of ergonomics studies include "use", "work", and "task". In addition, the t-SNE technique revealed that words, such as "workplace", "design", and "engineering," exhibited the highest relevance to ergonomics. The keywords observed in the abstract of ergonomic studies using t-SNE were classified into four groups. Ergonomics studies registered with MEDLINE have investigated the risk factors associated with workers performing an operation or task using tools, and in this study, ergonomics studies were identified by the relationship between keywords using word embedding. The results of this study will provide useful and diverse insights on future research direction on ergonomic studies.

• Journal of Platform Technology
• /
• v.8 no.4
• /
• pp.38-46
• /
• 2020

This study describes how to build a security control system using an open source big data solution so that private companies can build an overall security control infrastructure. In particular, the infrastructure was built using the Elastic Stack, one of the free open source big data analysis solutions, as a way to shorten the cost and development time when building a security control system. A comparative experiment was conducted. In addition, as a result of comparing and analyzing the functions, convenience, service and technical support of the two solution, it was found that the Elastic Stack has advantages in the security control of Big Data in terms of community and open solution. Using the Elastic Stack, security logs were collected, analyzed, and visualized step by step to create a dashboard, input large logs, and measure the search speed. Through this, we discovered the possibility of the Elastic Stack as a big data analysis solution that could replace Splunk.

• Fashion & Textile Research Journal
• /
• v.23 no.6
• /
• pp.799-809
• /
• 2021

This study aimed to grasp the perceptions and trends in fashion platforms and fashion smart factories using big data analysis. As a research method, big data analysis, fashion platform, and smart factory were identified through literature and prior studies, and text mining analysis and network analysis were performed after collecting text from the web environment between April 2019 and April 2021. After data purification with Textom, the words of fashion platform (1,0591 pieces) and fashion smart factory (9750 pieces) were used for analysis. Key words were derived, the frequency of appearance was calculated, and the results were visualized in word cloud and N-gram. The top 70 words by frequency of appearance were used to generate a matrix, structural equivalence analysis was performed, and the results were displayed using network visualization and dendrograms. The collected data revealed that smart factory had high social issues, but consumer interest and academic research were insufficient, and the amount and frequency of related words on the fashion platform were both high. As a result of structural equalization analysis, it was found that fashion platforms with strong connectivity between clusters are creating new competitiveness with service platforms that add sharing, manufacturing, and curation functions, and fashion smart factories can expect future value to grow together, according to digital technology innovation and platforms. This study can serve as a foundation for future research topics related to fashion platforms and smart factories.

• Journal of the Institute of Convergence Signal Processing
• /
• v.22 no.4
• /
• pp.199-205
• /
• 2021

The distributing board in directly installed on the ground or the bottom surface of the building, and when vibrations such as earthquakes or external shocks occur, the possibility of damage or malfunction of electric components such as internal power devices, wiring, and protection relays increases. Recently, the need for a seismic type distributing board is increasing, and research and development of a distributing board having a vibration damping function and product launch are being conducted. In this paper, an IoT-based data collection device system capable of measuring vibration and environmental data of distributing board was designed and implemented. When vibration occurred on the distributing board, data was stored and visualized in the MySQL DB through Node-RED for monitoring and data storage using the MQTT protocol for reliable messaging transmission. The test was conducted by attaching the IoT device of the distributing board, and data was collected in real-time and monitored through Node-RED.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.45 no.2
• /
• pp.65-76
• /
• 2022

Recently, not only traditional statistical techniques but also machine learning algorithms have been used to make more accurate bankruptcy predictions. But the insolvency rate of companies dealing with financial institutions is very low, resulting in a data imbalance problem. In particular, since data imbalance negatively affects the performance of artificial intelligence models, it is necessary to first perform the data imbalance process. In additional, as artificial intelligence algorithms are advanced for precise decision-making, regulatory pressure related to securing transparency of Artificial Intelligence models is gradually increasing, such as mandating the installation of explanation functions for Artificial Intelligence models. Therefore, this study aims to present guidelines for eXplainable Artificial Intelligence-based corporate bankruptcy prediction methodology applying SMOTE techniques and LIME algorithms to solve a data imbalance problem and model transparency problem in predicting corporate bankruptcy. The implications of this study are as follows. First, it was confirmed that SMOTE can effectively solve the data imbalance issue, a problem that can be easily overlooked in predicting corporate bankruptcy. Second, through the LIME algorithm, the basis for predicting bankruptcy of the machine learning model was visualized, and derive improvement priorities of financial variables that increase the possibility of bankruptcy of companies. Third, the scope of application of the algorithm in future research was expanded by confirming the possibility of using SMOTE and LIME through case application.

• Journal of KIISE:Software and Applications
• /
• v.34 no.9
• /
• pp.828-847
• /
• 2007

In this paper we propose a formal data model of an object-based thesaurus system supporting multi-dimensional facets. According to facets reflecting on respective user perspectives, it supports systematic construction, browsing, navigating and referencing of thesauri. Unlike other faceted thesaurus systems, it systematically manages its complexity by appropriately ing sophisticated conceptual structure through visualized browsing and navigation as well as construction. The browsing and navigation is performed by dynamically generating multi-dimensional virtual thesaurus hierarchies called "faceted thesaurus hierarchies." The hierarchies are automatically constructed by combining facets, each representing a dimension of view. Such automatic construction may make it possible the flexible extension of thesauri for they can be easily upgraded by pure insertion or deletion of facets. With a well defined set of self-referential queries, the thesauri can also be effectively referenced from multiple view points since they are structured by appropriately interpreting the semantics of instances based on facets. In this paper, we first formalize the underlying model and then implement its prototype to demonstrate its feasibility.

• 한국가시화정보학회:학술대회논문집
• /
• 2002.11a
• /
• pp.1-6
• /
• 2002

How is the flow in a rotating impeller. About 35 years have passed since one experimentalist rotating with the impeller. of a huge centrifugal blower made the flow measurements using a hot-wire anemometer (Fowler 1968). Optical measurement methods have great advantages over the intrusive methods especially for the flow measurement in a rotating impeller. One is the optical flow visualization (FV) technique (Senoo, et al., 1968) and the other is the application of laser velocimetry (LV) (Hah and Krain, 1990). Particle image velocimetries (PIVs) combine major features of both FV and LV, and are very attractive due to the feasibility of simultaneous and multi-points measurements (Hayami and Aramaki, 1999). A high-pressure-ratio transonic centrifugal compressor with a low-solidity cascade diffuser was tested in a closed loop with HFC134a gas at 18,000rpm (Hayami, 2000). Two kinds of measurement techniques by image processing were applied to visualize a flow in the compressor. One is a velocity field measurement at the inducer of the impeller using a PIV and the other is a pressure field measurement on the side wall of the cascade diffuser using a pressure sensitive paint (PSP) measurement technique. The PIV was successfully applied for visualization of an unsteady behavior of a shock wave based on the instantaneous velocity field measurement (Hayami, et al., 2002b) as well as a phase-averaged velocity vector field with a shock wave over one blade pitch (Hayami, et al., 2002a. b). A violent change in pressure was successfully visualized using a PSP measurement during a surge condition even though there are still some problems to be overcome (Hayami, et al., 2002c). Both PIV and PSP results are discussed in comparison with those of laser-2-focus (L2F) velocimetry and those of semiconductor pressure sensors. Experimental fluid dynamics (EFDs) are still growing up more and more both in hardware and in software. On the other hand, computational fluid dynamics (CFDs) are very attractive to understand the details of flow. A secondary flow on the side wall of the cascade diffuser was visualized based either steady or unsteady CFD calculations (Bonaiuti, et al.,2002). EFD and CFD methods will be combined to a hybrid method being complementary to each other. Measurement techniques by image processing as well as CFD calculations give a huge amount of data. Then, data mining technique will become more important to understand the flow mechanism both for EFD and CFD.

• Therapeutic Science for Rehabilitation
• /
• v.13 no.2
• /
• pp.53-67
• /
• 2024

Objective : This study aimed to identify common activities with similar participation levels between community-dwelling individuals in the United States (US) and South Korea (Study 1), and analyze the changes in activity participation patterns across the adult lifespan in South Korea (Study 2). Methods : We administered the online survey-based Activity Card Sort version 3 (ACS-3) to adults living in the US and South Korea. In Study 1, we computed the average participation level and visualized 100 activities of the ACS-3 from both the US and Korean samples. The average participation level across the four age groups in Study 2 was calculated and visualized to understand the changes in patterns of involvement across the four ACS-3 domains in a Korean sample. Results : In Study 1, data from 161 Americans and 163 Koreans were analyzed. Of the 100 activities, 48 (instrumental: 20; leisure: 13; fitness/health: 6; social: 9) demonstrated similar levels of participation between the two samples. In Study 2, data from 420 Koreans were analyzed and a tendency for decreased participation with age was found in all domains, except for the instrumental domain. Conclusion : Common daily activities may be used as a means of intervention across cultures in occupational therapy. Protective approaches and support are recommended to optimize older adults' participation in daily life.