• Journal of KIISE:Computing Practices and Letters
• /
• v.9 no.3
• /
• pp.322-331
• /
• 2003

As the complexity of a 3D game is increased by various factors of the game scenario, it has a problem for controlling the interrelation of the game objects. Therefore, a game system has a necessity of the coordination of the responses of the game objects. Also, it is necessary to control the behaviors of animations of the game objects in terms of the game scenario. To produce realistic game simulations, a system has to include a structure for designing the interactions among the game objects. This paper presents a method that designs the dynamic control mechanism for the interaction of the game objects in the game scenario. For the method, we suggest a game agent system as a framework that is based on intelligent agents who can make decisions using specific rules. Game agent systems are used in order to manage environment data, to simulate the game objects, to control interactions among game objects, and to support visual authoring interface that ran define a various interrelations of the game objects. These techniques can process the autonomy level of the game objects and the associated collision avoidance method, etc. Also, it is possible to make the coherent decision-making ability of the game objects about a change of the scene. In this paper, the rule-based behavior control was designed to guide the simulation of the game objects. The rules are pre-defined by the user using visual interface for designing their interaction. The Agent State Decision Network, which is composed of the visual elements, is able to pass the information and infers the current state of the game objects. All of such methods can monitor and check a variation of motion state between game objects in real time. Finally, we present a validation of the control method together with a simple case-study example. In this paper, we design and implement the supervised classification systems for high resolution satellite images. The systems support various interfaces and statistical data of training samples so that we can select the most effective training data. In addition, the efficient extension of new classification algorithms and satellite image formats are applied easily through the modularized systems. The classifiers are considered the characteristics of spectral bands from the selected training data. They provide various supervised classification algorithms which include Parallelepiped, Minimum distance, Mahalanobis distance, Maximum likelihood and Fuzzy theory. We used IKONOS images for the input and verified the systems for the classification of high resolution satellite images.

• The Journal of Korean Academy of Sensory Integration
• /
• v.13 no.2
• /
• pp.63-73
• /
• 2015

Objective : To report the effects of a specific intervention, the Interactive Metronome$^{(R)}$ (IM), on timing, attention and motor function of a children with ADHD. Methods : The study is case reports about two boys with ADHD. One boy who is born 2008 is attending general elementary school as a first year student (case 1), and another boy who is born 2001 is attending general elementary school as a second year student (case 2). For each case subject, IM training was provided during 3 weeks, from January 2015 to Febrary 2015. Evaluations were performed pre- and post-intervention in order to exam timing, attention and motor skills. The measurements uses in this study are Long Form Assessment (LFA) for the timing, RehaCom screening module for the attention, and Bruininks-Oseretsky Test of Morot Proficiency, second version (BOT-2) for the motor function. Results : The timing function was improved in both cases since both showed reduced response time for all motor tasks of LFA. In terms of attention, case 1 showed improvement of visual attention division, neglect and response Inhibition, and case 2 showed improvement of sustained attention. Lastly, in the BOT-2, case 1 showed improved the percentile rank of short (from 42%ile to 96%ile), and case 2 also showed similar improvement (from 21%ile to 66%ile). Conclusion : This study provides positive evidence that the Interactive Metronome$^{(R)}$ training has positive power to facilitate several body functions such as timing, attention and motor control of children with ADHD, through two case studies.

• Journal of Industrial Convergence
• /
• v.20 no.9
• /
• pp.59-70
• /
• 2022

Background: Diabetes education is an important factor in blood glucose control. Reinforced education is necessary for effective diabetes education. However, it is difficult to provide reinforced diabetes education within Korea's medical environment. Therefore, we want to analyze the effect of continuous diabetes education using mobile health care that can effectively provide repeated education without having to face the patient. Methods: This study is a multicenter, randomized, controlled, pre-post design study conducted to analyze the effect of a continuous diabetes education method. A total of 109 people were registered at five hospitals in south Korea, and they were randomly assigned to the app group (34 people) who received real-time coaching and repetitive training, the logbook group (37 people) who received face-to-face training after writing a blood glucose logbook, and the general group (38 people) who received a one-time diabetes education. The study was conducted for a total of 24 weeks. Twenty-one patients withdrew their consent and failed to perform an HbA1c. A final 88 patients were analyzed. The difference in HbA1c, Self-management behavior, and Quality of life before and after education was analyzed. Results: The study involved 51 (58%) male subjects, mean age was 55.8 years and mean duration of diabetes was 7.6 years. After 24 weeks of intervention, there was no significant difference in self-care behavior and quality of life between the three groups, but the HbA1c of the app group significantly decreased after education compared to the logbook group and the general group (F=4.62, p=.013). Conclusion: It can be seen through the app group that receiving real-time education is more effective in improving blood glucose management and continuous diabetes education is important.

• The Journal of Korean society of community based occupational therapy
• /
• v.6 no.1
• /
• pp.13-23
• /
• 2016

Objective : The aim of the present study was to investigate the effect of aloud reading on handwriting legibility in low-level elementary school children. Methods : The subject of the present study consisted of 45 elementary school 2nd graders who were normally developed checked by the Developmental Test of Visual Perception. Experimental period was conducted total six times that Pre-evaluation once, four times intervention, and Post-evaluation once from November 2 to November 25, 2015. When Pre-evaluation and Post-evaluation was measured the handwriting legibility and speed using Handwriting Skill Test. When intervention divided and implemented to experimental group who handwriting with aloud reading, control group1 who only handwriting, and control group2 who nothing. Pre-evaluation and Post-evaluation identified change the handwriting legibility using Handwriting Skill Test. Result : After intervention, handwriting legibility improve female than male. At word card1, control group1 improve significantly handwriting legibility within group and control group2 come out significant difference but handwriting legibility decrease. At word card2, experimental group and control group1 improve significantly handwriting legibility within group. Experimental group and control group1, control group1 and control group2 come out significant difference between group. Conclusion : The present study demonstrates that handwriting training improve handwriting legibility to elementary school 2nd graders.

• Korean Journal of Human Ecology
• /
• v.7 no.1
• /
• pp.51-67
• /
• 2004

This study analyzed the effects of nutrition technicians' hygiene education on cooking workers' performance of hygiene management in order to ensure the security of school meals. The situation of cooking workers' disposition in subject schools was elementary school(51.1%) and middle school(48.9%) and the type of meals was rural area type(54.2%), urban type(36.5%). and island and isolated area type(9.3%). The methods of meals management were single cooking(88.2%) and joint cooking and management(11.8%). The type of distributing meals was distributing in a dining room(93.5%), in a classroom(3.7%), and in both dining room and classroom(2.8%). Nutrition technicians' employment form included regular(53.5%) and daily(88.2%). Their education was junior college graduate(50.2%), university graduate(44.8%). and graduate school students(5.0%). Cooking workers' employment form included daily(88.2%) and regular (11.1%). suggesting that most were regular. Most cooking workers(77.4%) had at least high school certificate. Regarding the situation of cooking workers' disposition in subject schools, the number of student per one cooking worker was found as 91-120(37.2%), 61-90(22.6%). 60 and under(21.l %). 121-15006.7%). and 151 and over(2.5%). Cooking workers' level of performance of hygiene management was post-working stage(66.37/75 marks), pre-working stage(64.22/75 marks). and working stage(20.34/25 marks), The counting of meals articles in a pre-working stage(20.34/25 marks). temperature and required time in a working stage(18.78/25 marks), and machinery equipment and hygiene in a pre-working stage(21.40/25 marks) showed lowest of performance, which suggest poor service of hygiene. Cooking workers' performance of hygiene management by working stage showed the significant difference with school class(p<.001), type of schools with meals(p<.05). state of cooking workers' employment(p<.001), and cooking worker's disposition(p<.05). A working stage showed the significant difference with type of schools with meals(p<.05). A post-working stage showed the significant difference according to type of schools with meals(p<.05), and the methods of meals management(p<.05), and cooking workers' disposition(p<.05). In the execution of hygiene education, individual hygiene was highest(94.8%), followed by the management of machinery equipment and tools(89.7%), food poisoning and microorganism(94.7%), and the method of food treatment(76.4%). A yearly plan of hygiene education included established(83.9%) and not established(l6.1%). Regular education included not executed(25.1%), 2-3 times a month(l6.1%), and more than 4 a month(4.0%) and occasional education was not executed(57.0%), 1-3 times a month(26.3%), and more than 4 a month(l5.7%). In the methods for hygiene education, oral education(95.7%) was used most, followed by demonstration(10.5%), poster/photo(10.5%), video/slide(3.7%), and computer(3.7%). Frequency of improvement and complement of hygiene education included once a month(56.3%), once a year(20.7%), by quarter(l1.5%), and every six months(1l.5%). Newspaper was used most in materials of hygiene education, followed by internet, TV, nutrition technician's reeducation, information exchange between members, educational office's training, and reference book, and educational office's material. and symposium. Cooking workers' assessment of the effect of hygiene education was conducted through observation(56.8%), check table(l5.2%), question(l4.0%), and examination(14.0%). The reason of cooking workers' low level of performance included habitual custom(53.9%), lack of understanding(20.4%), overwork(l4.6%), and lack of knowledge(l1.l%) and the reason of difficulty in hygiene education included lack of time(55.3%), lack of understanding(27.6%), lack of knowledge and information(8.7%), and lack of budget(48.0%).

• KIPS Transactions on Software and Data Engineering
• /
• v.12 no.11
• /
• pp.471-480
• /
• 2023

The Korean construction order volume in South Korea grew significantly from 91.3 trillion won in public orders in 2013 to a total of 212 trillion won in 2021, particularly in the private sector. As the size of the domestic and overseas markets grew, the scale and complexity of EPC (Engineering, Procurement, Construction) projects increased, and risk management of project management and ITB (Invitation to Bid) documents became a critical issue. The time granted to actual construction companies in the bidding process following the EPC project award is not only limited, but also extremely challenging to review all the risk terms in the ITB document due to manpower and cost issues. Previous research attempted to categorize the risk terms in EPC contract documents and detect them based on AI, but there were limitations to practical use due to problems related to data, such as the limit of labeled data utilization and class imbalance. Therefore, this study aims to develop an AI model that can categorize the contract terms based on the FIDIC Yellow 2017(Federation Internationale Des Ingenieurs-Conseils Contract terms) standard in detail, rather than defining and classifying risk terms like previous research. A multi-text classification function is necessary because the contract terms that need to be reviewed in detail may vary depending on the scale and type of the project. To enhance the performance of the multi-text classification model, we developed the ELECTRA PLM (Pre-trained Language Model) capable of efficiently learning the context of text data from the pre-training stage, and conducted a four-step experiment to validate the performance of the model. As a result, the ensemble version of the self-developed ITB-ELECTRA model and Legal-BERT achieved the best performance with a weighted average F1-Score of 76% in the classification of 57 contract terms.

• Journal of Intelligence and Information Systems
• /
• v.24 no.1
• /
• pp.205-225
• /
• 2018

Convolutional Neural Network (ConvNet) is one class of the powerful Deep Neural Network that can analyze and learn hierarchies of visual features. Originally, first neural network (Neocognitron) was introduced in the 80s. At that time, the neural network was not broadly used in both industry and academic field by cause of large-scale dataset shortage and low computational power. However, after a few decades later in 2012, Krizhevsky made a breakthrough on ILSVRC-12 visual recognition competition using Convolutional Neural Network. That breakthrough revived people interest in the neural network. The success of Convolutional Neural Network is achieved with two main factors. First of them is the emergence of advanced hardware (GPUs) for sufficient parallel computation. Second is the availability of large-scale datasets such as ImageNet (ILSVRC) dataset for training. Unfortunately, many new domains are bottlenecked by these factors. For most domains, it is difficult and requires lots of effort to gather large-scale dataset to train a ConvNet. Moreover, even if we have a large-scale dataset, training ConvNet from scratch is required expensive resource and time-consuming. These two obstacles can be solved by using transfer learning. Transfer learning is a method for transferring the knowledge from a source domain to new domain. There are two major Transfer learning cases. First one is ConvNet as fixed feature extractor, and the second one is Fine-tune the ConvNet on a new dataset. In the first case, using pre-trained ConvNet (such as on ImageNet) to compute feed-forward activations of the image into the ConvNet and extract activation features from specific layers. In the second case, replacing and retraining the ConvNet classifier on the new dataset, then fine-tune the weights of the pre-trained network with the backpropagation. In this paper, we focus on using multiple ConvNet layers as a fixed feature extractor only. However, applying features with high dimensional complexity that is directly extracted from multiple ConvNet layers is still a challenging problem. We observe that features extracted from multiple ConvNet layers address the different characteristics of the image which means better representation could be obtained by finding the optimal combination of multiple ConvNet layers. Based on that observation, we propose to employ multiple ConvNet layer representations for transfer learning instead of a single ConvNet layer representation. Overall, our primary pipeline has three steps. Firstly, images from target task are given as input to ConvNet, then that image will be feed-forwarded into pre-trained AlexNet, and the activation features from three fully connected convolutional layers are extracted. Secondly, activation features of three ConvNet layers are concatenated to obtain multiple ConvNet layers representation because it will gain more information about an image. When three fully connected layer features concatenated, the occurring image representation would have 9192 (4096+4096+1000) dimension features. However, features extracted from multiple ConvNet layers are redundant and noisy since they are extracted from the same ConvNet. Thus, a third step, we will use Principal Component Analysis (PCA) to select salient features before the training phase. When salient features are obtained, the classifier can classify image more accurately, and the performance of transfer learning can be improved. To evaluate proposed method, experiments are conducted in three standard datasets (Caltech-256, VOC07, and SUN397) to compare multiple ConvNet layer representations against single ConvNet layer representation by using PCA for feature selection and dimension reduction. Our experiments demonstrated the importance of feature selection for multiple ConvNet layer representation. Moreover, our proposed approach achieved 75.6% accuracy compared to 73.9% accuracy achieved by FC7 layer on the Caltech-256 dataset, 73.1% accuracy compared to 69.2% accuracy achieved by FC8 layer on the VOC07 dataset, 52.2% accuracy compared to 48.7% accuracy achieved by FC7 layer on the SUN397 dataset. We also showed that our proposed approach achieved superior performance, 2.8%, 2.1% and 3.1% accuracy improvement on Caltech-256, VOC07, and SUN397 dataset respectively compare to existing work.

• Journal of Intelligence and Information Systems
• /
• v.26 no.2
• /
• pp.79-104
• /
• 2020

Recently, as deep learning has attracted attention, the use of deep learning is being considered as a method for solving problems in various fields. In particular, deep learning is known to have excellent performance when applied to applying unstructured data such as text, sound and images, and many studies have proven its effectiveness. Owing to the remarkable development of text and image deep learning technology, interests in image captioning technology and its application is rapidly increasing. Image captioning is a technique that automatically generates relevant captions for a given image by handling both image comprehension and text generation simultaneously. In spite of the high entry barrier of image captioning that analysts should be able to process both image and text data, image captioning has established itself as one of the key fields in the A.I. research owing to its various applicability. In addition, many researches have been conducted to improve the performance of image captioning in various aspects. Recent researches attempt to create advanced captions that can not only describe an image accurately, but also convey the information contained in the image more sophisticatedly. Despite many recent efforts to improve the performance of image captioning, it is difficult to find any researches to interpret images from the perspective of domain experts in each field not from the perspective of the general public. Even for the same image, the part of interests may differ according to the professional field of the person who has encountered the image. Moreover, the way of interpreting and expressing the image also differs according to the level of expertise. The public tends to recognize the image from a holistic and general perspective, that is, from the perspective of identifying the image's constituent objects and their relationships. On the contrary, the domain experts tend to recognize the image by focusing on some specific elements necessary to interpret the given image based on their expertise. It implies that meaningful parts of an image are mutually different depending on viewers' perspective even for the same image. So, image captioning needs to implement this phenomenon. Therefore, in this study, we propose a method to generate captions specialized in each domain for the image by utilizing the expertise of experts in the corresponding domain. Specifically, after performing pre-training on a large amount of general data, the expertise in the field is transplanted through transfer-learning with a small amount of expertise data. However, simple adaption of transfer learning using expertise data may invoke another type of problems. Simultaneous learning with captions of various characteristics may invoke so-called 'inter-observation interference' problem, which make it difficult to perform pure learning of each characteristic point of view. For learning with vast amount of data, most of this interference is self-purified and has little impact on learning results. On the contrary, in the case of fine-tuning where learning is performed on a small amount of data, the impact of such interference on learning can be relatively large. To solve this problem, therefore, we propose a novel 'Character-Independent Transfer-learning' that performs transfer learning independently for each character. In order to confirm the feasibility of the proposed methodology, we performed experiments utilizing the results of pre-training on MSCOCO dataset which is comprised of 120,000 images and about 600,000 general captions. Additionally, according to the advice of an art therapist, about 300 pairs of 'image / expertise captions' were created, and the data was used for the experiments of expertise transplantation. As a result of the experiment, it was confirmed that the caption generated according to the proposed methodology generates captions from the perspective of implanted expertise whereas the caption generated through learning on general data contains a number of contents irrelevant to expertise interpretation. In this paper, we propose a novel approach of specialized image interpretation. To achieve this goal, we present a method to use transfer learning and generate captions specialized in the specific domain. In the future, by applying the proposed methodology to expertise transplant in various fields, we expected that many researches will be actively conducted to solve the problem of lack of expertise data and to improve performance of image captioning.

• Journal of Korean Academy of Nursing Administration
• /
• v.6 no.1
• /
• pp.55-81
• /
• 2000

This study examined nursing care in a Newborn Intensive Care Unit (NICU) by reviewing nursing activities for the newborns. Through direct observation, time used for nursing care according to the nursing activity, shift, day of the week, and position of the nurses was measured. This study was done on November 15, 21, 24, 1999 at a university medical center hospital and included eight nurses and 179 newborns as the study subjects. The data were collected from the medical records, and by using a nursing activity record for the NICU, and a nursing activity time record for the NICU. The first step in the data collection process was to develop a list of nursing activities which was done through a literature review, examination of medical affairs and duty records. Content validity was measured by a panel of three professors who were experienced clinicians. In the second step two pre-training sessions were held with three sophomore student nurses who then measured the time for each nursing activity using a stopwatch. The data were analyzed using frequencies for nursing activities, averages, percentages and ANOVA for differences between shift and between days of the week, and percentages and t-test for differences according to position of the nurse. The results are as follows: 1) The total number of activities was 156, direct or indirect nursing activities. Direct nursing activity classified according to physical, educational, emotional/social/economic/spiritual needs. There were 109 direct nursing activities in 16 fields. 2) The order of nursing activities, according to time required, was record keeping, nutritional care, measurement/observation, medication, hygiene care, examination and specimen collection, and checking supplies, and according to frequency, measurement/ examination, record keeping, nutrition care, hygiene care, elimination care and medication. 3) According to shift, direct care during the night shift at 313.4 minutes was the longest time and indirect nursing care during the night shift at 252.2 minutes was the highest time. 4) For days of the week, Monday had the highest time for direct care 275.8 minutes (34.6%) and Wednessday had the highest time for indirect nursing care 269.6 minutes (36.1%). 5) For nursing time according to position of nurse, general nurses had the highest for direct care (330.7 minutes), nurse managers for indirect nursing activities (239.0 minutes) and general nurses for individual private time (63.9 minutes). The results of this study show that the major nursing time consuming activities included record keeping, nutrition care and measurement/examination. For newborns, time needs to be allowed for care to be sensitive, sophisticated and specialized rather than concentrated on indirect nursing tasks such as record keeping. Therefore, it is imperative to develop computerized systems that support a systematic approach to record keeping which is more efficient. Moreover, nursing needs according to shift, day or position of nurse can be utilized in assessing nursing resources through a computerized process.

• Journal of the Korea Society of Computer and Information
• /
• v.17 no.8
• /
• pp.189-201
• /
• 2012

The purpose of this study is to develop an effective teaching model for the "Problem solving methods and procedures" section in the revised academic high school informatics curriculum, verify its effectiveness, make the subject more effective and appealing to teachers as well as students. The model includes a middle school level informatics curriculum for the students who have yet to learn the section. This development follows the ADDIE model, and the Python programming language is adopted for the model. Using the model, classes were conducted with two groups: high school computer club students and undergraduate students majoring in computer education. Of the undergraduate students 75% responded positively to the model. This model was applied in the actual high school classroom teaching for 23 class-hours in the spring semester 2012. The Pearson correlation coefficient that verifies the correspondence between the PSI score and the informatics midterm exam grade is .247, which reflects a weak positive correlation. The result of the study showed that the developed teaching model is an effective tool in educating students about the "problem solving methods and procedures". The model is to be a cornerstone of teaching/learning plans for informatics at academic high school as well as training materials for pre-service teachers.