• Journal of Intelligence and Information Systems
• /
• v.27 no.1
• /
• pp.177-190
• /
• 2021

Due to the development of the fourth industrial revolution technology, efforts are being made to improve areas that humans cannot handle by utilizing artificial intelligence techniques such as machine learning. Although on-demand production companies also want to reduce corporate risks such as delays in delivery by predicting total production time for orders, they are having difficulty predicting this because the total production time is all different for each order. The Theory of Constraints (TOC) theory was developed to find the least efficient areas to increase order throughput and reduce order total cost, but failed to provide a forecast of total production time. Order production varies from order to order due to various customer needs, so the total production time of individual orders can be measured postmortem, but it is difficult to predict in advance. The total measured production time of existing orders is also different, which has limitations that cannot be used as standard time. As a result, experienced managers rely on persimmons rather than on the use of the system, while inexperienced managers use simple management indicators (e.g., 60 days total production time for raw materials, 90 days total production time for steel plates, etc.). Too fast work instructions based on imperfections or indicators cause congestion, which leads to productivity degradation, and too late leads to increased production costs or failure to meet delivery dates due to emergency processing. Failure to meet the deadline will result in compensation for delayed compensation or adversely affect business and collection sectors. In this study, to address these problems, an entity that operates an order production system seeks to find a machine learning model that estimates the total production time of new orders. It uses orders, production, and process performance for materials used for machine learning. We compared and analyzed OLS, GLM Gamma, Extra Trees, and Random Forest algorithms as the best algorithms for estimating total production time and present the results.

• Journal of Intelligence and Information Systems
• /
• v.28 no.2
• /
• pp.127-146
• /
• 2022

Recently, as word embedding has shown excellent performance in various tasks of deep learning-based natural language processing, researches on the advancement and application of word, sentence, and document embedding are being actively conducted. Among them, cross-language transfer, which enables semantic exchange between different languages, is growing simultaneously with the development of embedding models. Academia's interests in vector alignment are growing with the expectation that it can be applied to various embedding-based analysis. In particular, vector alignment is expected to be applied to mapping between specialized domains and generalized domains. In other words, it is expected that it will be possible to map the vocabulary of specialized fields such as R&D, medicine, and law into the space of the pre-trained language model learned with huge volume of general-purpose documents, or provide a clue for mapping vocabulary between mutually different specialized fields. However, since linear-based vector alignment which has been mainly studied in academia basically assumes statistical linearity, it tends to simplify the vector space. This essentially assumes that different types of vector spaces are geometrically similar, which yields a limitation that it causes inevitable distortion in the alignment process. To overcome this limitation, we propose a deep learning-based vector alignment methodology that effectively learns the nonlinearity of data. The proposed methodology consists of sequential learning of a skip-connected autoencoder and a regression model to align the specialized word embedding expressed in each space to the general embedding space. Finally, through the inference of the two trained models, the specialized vocabulary can be aligned in the general space. To verify the performance of the proposed methodology, an experiment was performed on a total of 77,578 documents in the field of 'health care' among national R&D tasks performed from 2011 to 2020. As a result, it was confirmed that the proposed methodology showed superior performance in terms of cosine similarity compared to the existing linear vector alignment.

• Journal of fish pathology
• /
• v.35 no.1
• /
• pp.1-25
• /
• 2022

The aquaculture industry has developed rapidly over the last three decades and is an important industry that supplies over 15% of humans' animal protein intake; therefore, there is a need to increase production to meet the continuous demand. The fish cage farms on the southern coast (Kyengsangnam-do and Jeollanam-do) of Korea are critical resources in aquaculture because they account for approximately 90% of the national total fish cage farms by water area ratio. However, the current aquaculture environment is being gradually affected by climate change, which is a global issue, and its effects are expected to intensify in the future. Therefore, it is urgently imperative to accurately evaluate the effects of climate change on South Korean aquaculture industries and to develop social and national strategies to minimize damage to the fishing industry. The damage to fish farmed in cage farms on the southern coast is increasing annually and the leading causes are high and low water temperature and red tides, which are directly or indirectly related to climate change. At present, global warming can provide opportunities for aquaculture industrialization of fish or other novel species, with economic implications. However, despite such opportunities, the influx of new species can also cause problems such as ecological disturbances, increase in the reproduction frequency of microalgae such as red tide, increase in disease incidence, and occurrence and periods of high water temperatures in summer. The scale of farmed fish mortality is increasing due to the complex effects of these factors. Increased damages due to fish mortality not only have severe economic impacts on the aquaculture industry, but the social costs of responding to the damage and follow-up measures also increase. various active responses can reduce the mortality damage in fish farms such as improving the management skills in aquaculture, improved species breeding, efficient food management, disease prevention, proactive responses, and system-wide improvements. This review article analyzes the large-scale mortality cases occurring in fish cage farms on the southern coast of Korea and proposes measures to mitigate mortality and enhance responses to such scenarios.

• Food Science and Preservation
• /
• v.31 no.3
• /
• pp.423-432
• /
• 2024

In this study, a vacuum tumbler with 4 impellers (DVT) was designed and applied for thawing frozen pork (vacuum -60 kPa, jacket 35℃, 1 rpm). Quality characteristics of the thawed pork were compared with those of industrially thawed meat by natural air at room temperature (NAT) and imported vacuum tumbler (IVT). The thawing time for frozen pork (303.36 kg) using DVT (165 min) was much shorter than that of NAT (4,200 min). DVT-thawed pork had lower drip loss (0.85%) than NAT (2.08%). DVT-thawed pork showed a pH of 5.92, a total bacterial count of 1.96±0.02 log CFU/g and no coliforms. Deteriorations in fat (TBARS 0.31±0.01 MDA mg/kg) and protein (VBN 5.67±1.98 mg%) in DVT-thawed pork were significantly lower than those of NAT (p<0.05). DVT-thawed pork had a high water-holding capacity (WHC, 97.5%). The hardness (34.59±0.46 N) and chewiness (188.21±0.17) of cooked DVT-thawed pork were about 5-6 times lower than those of NTA. Microstructure (SEM) showed myofibrillar damage in NAT-thawed pork, whereas dense myofibrillar structure was observed in DVT-thawed pork. DVT was better or similar to IVT in all evaluation parameters. The designed DVT is expected to be used as an efficient thawing method in terms of processing time and yield and to produce thawed meat with high WHC, soft texture, and low spoilage by minimizing tissue damage.

• Journal of agricultural medicine and community health
• /
• v.25 no.2
• /
• pp.353-377
• /
• 2000

This study was performed to investigate the performance state and improvement countermeasure of Primary Health care Posts(PHPs). The operation reports of PHPs(1996 330 PHPs, 1999 313 PHPs) located in Kyongsangbuk-Do and data collected by self-administered questionnaire survey of 280 community health practitioners(CHPs) were analyzed. The major results were as follows: Population per PHP in 1999 decreased in number compared with 1996. But population of the aged increased in number. The performance status of PHP in 1999 increased compared with 1996. A hundred forty one community health practitioners(50.4%) replied that the fiscal standing of PHP was good. Only 1.4% replied that the fiscal standing of PHP was difficult. For the degree of satisfaction in affairs, overall of community health practitioners felt proud. The degree of cooperation between PHP and public health institutions was high and the degree of cooperation of between PHP and private medical institutions was high. The degree of cooperation between PHP and Health Center was significantly different by age of CHP, the service period of CHP, and CHP's service period at present PHP. Over seventy percent of CHPs replied that they had cooperative relationship with operation council, village health workers, community organization. CHPs who drew up the paper on PHP's health activity plan were 96.4 % and only 11.4% of CHPs participated drawing up the report on the second community health plan. CHPs who grasped the blood pressure and smoking status of residents over 70% were 88.2%, 63.9% respectively and the grasp rate of blood pressure fur residents were significantly different according to age and educational level of CHP. CHPs received job education in addition continuous job education arid participated on research program in last 3 years were 27.5%, respectively. CHPs performed the return health program for residents in last 3years were 65.4%. Over 95% of CHPs replied that PHPs might be necessary and 53.9% of CHPs replied that the role of PHPs should be increased. CHPS indicated that major reasons of FHPs lockout were lack of understanding for PHP and administrative convenience, CHPs were officials in special government service governors intention of self-governing body. CHPs suggested number of population in health need such as the aged and patients with chronic disease, opinion of residents, population size, traffic situation and network in order as evaluation criteria for PHP and suggested results of health performance, degree of relationship with residents, results of medical examination anti treatment, ability for administration and affairs in order as evaluation criteria for CHP. CHPs replied that the important countermeasures for PHPs under standard were affairs improvement of PHPs and shifting of location to health weakness area in city. Over 50% of CHPs indicated that the most important thing for improvement of PHPs was affairs adjustment of CLIP. And CHPs suggested that health programs carried out in priority at PHP were management of diabetes mellitus and hypertention. home visiting health care, health care for the aged. The Affairs of BLIP should be adjusted to satisfy community health need and health programs such as management of diabetes mellitus and hypertention, home visiting health care, health care for the aged should be activated in order that PHPs become organization reflecting value system of primary health care.

• Journal of Intelligence and Information Systems
• /
• v.24 no.1
• /
• pp.1-23
• /
• 2018

From the 21st century, various high-quality services have come up with the growth of the internet or 'Information and Communication Technologies'. Especially, the scale of E-commerce industry in which Amazon and E-bay are standing out is exploding in a large way. As E-commerce grows, Customers could get what they want to buy easily while comparing various products because more products have been registered at online shopping malls. However, a problem has arisen with the growth of E-commerce. As too many products have been registered, it has become difficult for customers to search what they really need in the flood of products. When customers search for desired products with a generalized keyword, too many products have come out as a result. On the contrary, few products have been searched if customers type in details of products because concrete product-attributes have been registered rarely. In this situation, recognizing texts in images automatically with a machine can be a solution. Because bulk of product details are written in catalogs as image format, most of product information are not searched with text inputs in the current text-based searching system. It means if information in images can be converted to text format, customers can search products with product-details, which make them shop more conveniently. There are various existing OCR(Optical Character Recognition) programs which can recognize texts in images. But existing OCR programs are hard to be applied to catalog because they have problems in recognizing texts in certain circumstances, like texts are not big enough or fonts are not consistent. Therefore, this research suggests the way to recognize keywords in catalog with the Deep Learning algorithm which is state of the art in image-recognition area from 2010s. Single Shot Multibox Detector(SSD), which is a credited model for object-detection performance, can be used with structures re-designed to take into account the difference of text from object. But there is an issue that SSD model needs a lot of labeled-train data to be trained, because of the characteristic of deep learning algorithms, that it should be trained by supervised-learning. To collect data, we can try labelling location and classification information to texts in catalog manually. But if data are collected manually, many problems would come up. Some keywords would be missed because human can make mistakes while labelling train data. And it becomes too time-consuming to collect train data considering the scale of data needed or costly if a lot of workers are hired to shorten the time. Furthermore, if some specific keywords are needed to be trained, searching images that have the words would be difficult, as well. To solve the data issue, this research developed a program which create train data automatically. This program can make images which have various keywords and pictures like catalog and save location-information of keywords at the same time. With this program, not only data can be collected efficiently, but also the performance of SSD model becomes better. The SSD model recorded 81.99% of recognition rate with 20,000 data created by the program. Moreover, this research had an efficiency test of SSD model according to data differences to analyze what feature of data exert influence upon the performance of recognizing texts in images. As a result, it is figured out that the number of labeled keywords, the addition of overlapped keyword label, the existence of keywords that is not labeled, the spaces among keywords and the differences of background images are related to the performance of SSD model. This test can lead performance improvement of SSD model or other text-recognizing machine based on deep learning algorithm with high-quality data. SSD model which is re-designed to recognize texts in images and the program developed for creating train data are expected to contribute to improvement of searching system in E-commerce. Suppliers can put less time to register keywords for products and customers can search products with product-details which is written on the catalog.

• Journal of Radiation Protection and Research
• /
• v.32 no.3
• /
• pp.105-110
• /
• 2007

Purpose : In spite of recent remarkable improvement of diagnostic imaging modalities such as CT, MRI, and PET and radiation therapy planing systems, ICR plan of uterine cervix cancer, based on recommendation of ICRU38(2D film-based) such as Point A, is still used widely. A 3-dimensional ICR plan based on CT image provides dose-volume histogram(DVH) information of the tumor and normal tissue. In this study, we compared tumor-dose, rectal-dose and bladder-dose through an analysis of DVH between CTV plan and ICRU38 plan based on CT image. Method and Material : We analyzed 11 patients with a cervix cancer who received the ICR of Ir-192 HDR. After 40Gy of external beam radiation therapy, ICR plan was established using PLATO(Nucletron) v.14.2 planing system. CT scan was done to all the patients using CT-simulator(Ultra Z, Philips). We contoured CTV, rectum and bladder on the CT image and established CTV plan which delivers the 100% dose to CTV and ICRU plan which delivers the 100% dose to the point A. Result : The volume$(average{\pm}SD)$ of CTV, rectum and bladder in all of 11 patients is $21.8{\pm}6.6cm^3,\;60.9{\pm}25.0cm^3,\;111.6{\pm}40.1cm^3$ respectively. The volume covered by 100% isodose curve is $126.7{\pm}18.9cm^3$ in ICRU plan and $98.2{\pm}74.5cm^3$ in CTV plan(p=0.0001), respectively. In (On) ICRU planning, $22.0cm^3$ of CTV volume was not covered by 100% isodose curve in one patient whose residual tumor size is greater than 4cm, while more than 100% dose was irradiated unnecessarily to the normal organ of $62.2{\pm}4.8cm^3$ other than the tumor in the remaining 10 patients with a residual tumor less than 4cm in size. Bladder dose recommended by ICRU 38 was $90.1{\pm}21.3%$ and $68.7{\pm}26.6%$ in ICRU plan and in CTV plan respectively(p=0.001) while rectal dose recommended by ICRU 38 was $86.4{\pm}18.3%$ and $76.9{\pm}15.6%$ in ICRU plan and in CTV plan, respectively(p=0.08). Bladder and rectum maximum dose was $137.2{\pm}50.1%,\;101.1{\pm}41.8%$ in ICRU plan and $107.6{\pm}47.9%,\;86.9{\pm}30.8%$ in CTV plan, respectively. Therefore, the radiation dose to normal organ was lower in CTV plan than in ICRU plan. But the normal tissue dose was remarkably higher than a recommended dose in CTV plan in one patient whose residual tumor size was greater than 4cm. The volume of rectum receiving more than 80% isodose (V80rec) was $1.8{\pm}2.4cm^3$ in ICRU plan and $0.7{\pm}1.0cm^3$ in CTV plan(p=0.02). The volume of bladder receiving more than 80% isodose(V80bla) was $12.2{\pm}8.9cm^3$ in ICRU plan and $3.5{\pm}4.1cm^3$ in CTV plan(p=0.005). According to these parameters, CTV plan could also save more normal tissue compared to ICRU38 plan. Conclusion : An unnecessary excessive radiation dose is irradiated to normal tissues within 100% isodose area in the traditional ICRU plan in case of a small size of cervix cancer, but if we use CTV plan based on CT image, the normal tissue dose could be reduced remarkably without a compromise of tumor dose. However, in a large tumor case, we need more research on an effective 3D-planing to reduce the normal tissue dose.