• Tunnel and Underground Space
• /
• v.30 no.5
• /
• pp.462-472
• /
• 2020

Rock classification is fundamental discipline of exploring geological and geotechnical features in a site, which, however, may not be easy works because of high diversity of rock shape and color according to its origin, geological history and so on. With the great success of convolutional neural networks (CNN) in many different image-based classification tasks, there has been increasing interest in taking advantage of CNN to classify geological material. In this study, a feasibility of the deep CNN is investigated for automatically and accurately identifying rock types, focusing on the condition of various shapes and colors even in the same rock type. It can be further developed to a mobile application for assisting geologist in classifying rocks in fieldwork. The structure of CNN model used in this study is based on a deep residual neural network (ResNet), which is an ultra-deep CNN using in object detection and classification. The proposed CNN was trained on 10 typical rock types with an overall accuracy of 84% on the test set. The result demonstrates that the proposed approach is not only able to classify rock type using images, but also represents an improvement as taking highly diverse rock image dataset as input.

• Journal of Chest Surgery
• /
• v.40 no.10
• /
• pp.680-684
• /
• 2007

Background: Complete surgical resection is the most effective treatment for stage IB non-small cell lung cancer (NSCLC). Recurrence accounts for the disappointing survival rates after resection. There has been renewed interest in adjuvant therapy after complete resection. Appropriate selection of effective adjuvant therapy will depend on the prognostic factors for recurrence. Material and Method: The study included 114 patients with completely resected stage IB NSCLC. The variables selected for the study were gender, age, the type of resection, cell type, the degree of differentiation, the tumor size and the presence of visceral pleura invasion. The Kaplan-Meier method was used to estimate the survival and disease-free survival rate. The results were compared using the log rank test. Multivariate analysis was performed by Cox's proportional hazard model. Two-sided p-valves < 0.05 were considered to be statistically significant. Result: The 3-year overall survival and the disease-free survival rates were 87.0% and 79.4%, respectively. The degree of differentiation showed a significant influence on disease-free survival according to the univariate analysis. According to the multivariate analysis, a poor grade of differentiation was a significant poor prognostic factor. Conclusion: These results demonstrate that poor differentiation may be a poor prognostic factor for patients with completely resected IB NSCLC. Therefore, the patients with a poor grade of differentiation may require adjuvant therapies.

• Journal of the Korea Society of Computer and Information
• /
• v.28 no.3
• /
• pp.189-196
• /
• 2023

This paper analyzes the level of satisfaction of two groups of teachers who were educated about artificial intelligence using App Inventor. The participants were 13 pre-service and 9 in-service elementary school teachers and the data was collected using a questionnaire. As a result of the study, in-service teachers were all more satisfied than pre-service teachers in terms of interest, difficulty, and participation in the education. In addition, the questions investigating whether education helped motivate learning of artificial intelligence and whether there is a willingness to apply it to elementary classes in the future were also more positive for in-service teachers than for pre-service teachers. In general, pre-service teachers had somewhat more negative views than in-service teachers, but they were more positive than in-service teachers in terms of whether the education helped improve their understanding of artificial intelligence and whether they were willing to participate in additional education. Analysis of the Mann-Whitney test to see if there was a significant difference in satisfaction between the two groups showed no significance. This may be because most of the students in the two groups already had block-type or text-type programming experience, so they were able to participate in the education without any special resistance or difficulty with App Inventor, resulting in high levels of satisfaction from both groups. The results of this study can provide basic data for the future development and operation of programs for artificial intelligence education for both pre-service and in-service elementary school teachers.

• Journal of Elementary Mathematics Education in Korea
• /
• v.13 no.2
• /
• pp.163-192
• /
• 2009

Currently, our country operates gifted education only as a special curriculum, which results in many problems, e.g., there are few beneficiaries of gifted education, considerable time and effort are required to gifted students, and gifted students' educational needs are ignored during the operation of regular curriculum. In order to solve these problems, the present study formulates the following research questions, finding it advisable to conduct gifted education in elementary regular classrooms within the scope of the regular curriculum. A. To devise a teaching plan for the gifted students on mathematics in the elementary school regular classroom. B. To develop a learning program for the gifted students in the elementary school regular classroom. C. To apply an in-depth learning program to gifted students in mathematics and analyze the effectiveness of the program. In order to answer these questions, a teaching plan was provided for the gifted students in mathematics using a differentiating instruction type. This type was developed by researching literature reviews. Primarily, those on characteristics of gifted students in mathematics and teaching-learning models for gifted education. In order to instruct the gifted students on mathematics in the regular classrooms, an in-depth learning program was developed. The gifted students were selected through teachers' recommendation and an advanced placement test. Furthermore, the effectiveness of the gifted education in mathematics and the possibility of the differentiating teaching type in the regular classrooms were determined. The analysis was applied through an in-depth learning program of selected gifted students in mathematics. To this end, an in-depth learning program developed in the present study was applied to 6 gifted students in mathematics in one first grade class of D Elementary School located in Nowon-gu, Seoul through a 10-period instruction. Thereafter, learning outputs, math diaries, teacher's checklist, interviews, video tape recordings the instruction were collected and analyzed. Based on instruction research and data analysis stated above, the following results were obtained. First, it was possible to implement the gifted education in mathematics using a differentiating instruction type in the regular classrooms, without incurring any significant difficulty to the teachers, the gifted students, and the non-gifted students. Specifically, this instruction was effective for the gifted students in mathematics. Since the gifted students have self-directed learning capability, the teacher can teach lessons to the gifted students individually or in a group, while teaching lessons to the non-gifted students. The teacher can take time to check the learning state of the gifted students and advise them, while the non-gifted students are solving their problems. Second, an in-depth learning program connected with the regular curriculum, was developed for the gifted students, and greatly effective to their development of mathematical thinking skills and creativity. The in-depth learning program held the interest of the gifted students and stimulated their mathematical thinking. It led to the creative learning results, and positively changed their attitude toward mathematics. Third, the gifted students with the most favorable results who took both teacher's recommendation and advanced placement test were more self-directed capable and task committed. They also showed favorable results of the in-depth learning program. Based on the foregoing study results, the conclusions are as follows: First, gifted education using a differentiating instruction type can be conducted for gifted students on mathematics in the elementary regular classrooms. This type of instruction conforms to the characteristics of the gifted students in mathematics and is greatly effective. Since the gifted students in mathematics have self-directed learning capabilities and task-commitment, their mathematical thinking skills and creativity were enhanced during individual exploration and learning through an in-depth learning program in a differentiating instruction. Second, when a differentiating instruction type is implemented, beneficiaries of gifted education will be enhanced. Gifted students and their parents' satisfaction with what their children are learning at school will increase. Teachers will have a better understanding of gifted education. Third, an in-depth learning program for gifted students on mathematics in the regular classrooms, should conform with an instructing and learning model for gifted education. This program should include various and creative contents by deepening the regular curriculum. Fourth, if an in-depth learning program is applied to the gifted students on mathematics in the regular classrooms, it can enhance their gifted abilities, change their attitude toward mathematics positively, and increase their creativity.

• Journal of Intelligence and Information Systems
• /
• v.26 no.2
• /
• pp.1-25
• /
• 2020

In this paper, we suggest an application system architecture which provides accurate, fast and efficient automatic gasometer reading function. The system captures gasometer image using mobile device camera, transmits the image to a cloud server on top of private LTE network, and analyzes the image to extract character information of device ID and gas usage amount by selective optical character recognition based on deep learning technology. In general, there are many types of character in an image and optical character recognition technology extracts all character information in an image. But some applications need to ignore non-of-interest types of character and only have to focus on some specific types of characters. For an example of the application, automatic gasometer reading system only need to extract device ID and gas usage amount character information from gasometer images to send bill to users. Non-of-interest character strings, such as device type, manufacturer, manufacturing date, specification and etc., are not valuable information to the application. Thus, the application have to analyze point of interest region and specific types of characters to extract valuable information only. We adopted CNN (Convolutional Neural Network) based object detection and CRNN (Convolutional Recurrent Neural Network) technology for selective optical character recognition which only analyze point of interest region for selective character information extraction. We build up 3 neural networks for the application system. The first is a convolutional neural network which detects point of interest region of gas usage amount and device ID information character strings, the second is another convolutional neural network which transforms spatial information of point of interest region to spatial sequential feature vectors, and the third is bi-directional long short term memory network which converts spatial sequential information to character strings using time-series analysis mapping from feature vectors to character strings. In this research, point of interest character strings are device ID and gas usage amount. Device ID consists of 12 arabic character strings and gas usage amount consists of 4 ~ 5 arabic character strings. All system components are implemented in Amazon Web Service Cloud with Intel Zeon E5-2686 v4 CPU and NVidia TESLA V100 GPU. The system architecture adopts master-lave processing structure for efficient and fast parallel processing coping with about 700,000 requests per day. Mobile device captures gasometer image and transmits to master process in AWS cloud. Master process runs on Intel Zeon CPU and pushes reading request from mobile device to an input queue with FIFO (First In First Out) structure. Slave process consists of 3 types of deep neural networks which conduct character recognition process and runs on NVidia GPU module. Slave process is always polling the input queue to get recognition request. If there are some requests from master process in the input queue, slave process converts the image in the input queue to device ID character string, gas usage amount character string and position information of the strings, returns the information to output queue, and switch to idle mode to poll the input queue. Master process gets final information form the output queue and delivers the information to the mobile device. We used total 27,120 gasometer images for training, validation and testing of 3 types of deep neural network. 22,985 images were used for training and validation, 4,135 images were used for testing. We randomly splitted 22,985 images with 8:2 ratio for training and validation respectively for each training epoch. 4,135 test image were categorized into 5 types (Normal, noise, reflex, scale and slant). Normal data is clean image data, noise means image with noise signal, relfex means image with light reflection in gasometer region, scale means images with small object size due to long-distance capturing and slant means images which is not horizontally flat. Final character string recognition accuracies for device ID and gas usage amount of normal data are 0.960 and 0.864 respectively.

• Journal of the Korea Convergence Society
• /
• v.10 no.7
• /
• pp.15-21
• /
• 2019

Research in detecting fake information gained a lot of interest after the US presidential election in 2016. Information from unknown sources are produced in the shape of news, and its rapid spread is fueled by the interest of public drawn to stimulating and interesting issues. In addition, the wide use of mass communication platforms such as social network services makes this phenomenon worse. Poynter Institute created the International Fact Checking Network (IFCN) to provide guidelines for judging the facts of skilled professionals and releasing "Code of Ethics" for fact check agencies. However, this type of approach is costly because of the large number of experts required to test authenticity of each article. Therefore, research in automated fake news detection technology that can efficiently identify it is gaining more attention. In this paper, we investigate fake news detection systems and researches that are rapidly developing, mainly thanks to recent advances in deep learning technology. In addition, we also organize shared tasks and training corpus that are released in various forms, so that researchers can easily participate in this field, which deserves a lot of research effort.

• Journal of Global Scholars of Marketing Science
• /
• v.9
• /
• pp.1-17
• /
• 2002

The purposes of this study were to investigate clothing involvement and clothing information source on personality types. Personality types were applied using Myers-Briggs Type Indicator(MBTI). Clothing involvement was measured using 17 questions of 5-point scales. Clothing information source was measured using 16 items of 5-point scales. The data were collected from 275 female using questionnaire. The data analysed with frequence, factor analysis, t-test, one-way ANOVA, and Duncan test. The clothing involvement included three factors: Symbolic Expression, Pleasure and Interest, and Fashionability. The clothing information source included four factors: Printed Media, Radio Media, Professional Media and Store Search. Extroversion-introvertion and thinking-feeling index had significant difference in the dimensions of clothing involvement. Extroversion types were significantly evaluated fashionability more than introvertion types(t=2.008, p<.05). Feeling types were significantly evaluated fashionability more than thinking types (t=2.428, p<.05). Extroversion-introvertion index had significant difference in clothing information source. Extroversion types were significantly used printed media more than Introvertion types.

• Proceedings of the International Microelectronics And Packaging Society Conference
• /
• 2004.09a
• /
• pp.233-241
• /
• 2004

A new alloy definition will be presented concerning increasing demands for the board level reliability of miniaturized interconnections. The damage mechanism for LFBGA components on different board finishes is not quite understood. Further demands from mobile phones are the drop test, characterizing interface performance of different package constructions in relation to decreased pad constructions and therefore interfaces. The paper discusses the characterization of interfaces based on SnPb, SnPbXYZ, SnAgCu and SnAgCuInNd ball materials and SnAgCuInNd as solder paste, the stability after accelerated tests and the description of modified interfaces strictly related to the assembly conditions, dissolution behavior of finishes on board side and the influence of intermetallic formation. The type of intermetallic as well as the quantity of intermetallics are observed, primaliry the hardness, E modules describing the ability of strain/stress compensation. First results of board level reliability are presented after TCT-40/+150. Improvement steps from the ball formulation will be discussed in conjunction to the implementation of lead free materials In order to optimize ball materials for area array devices accelareted aging conditions like TCTs were used to analyze the board level reliability of different ball materials for BGA, LFBGA, CSP, Flip Chip. The paper outlines lead-free ball analysis in comparison to conventional solder balls for BGA and chip size packages. The important points of interest are the description of processability related to existing ball attach procedures, requirements of interconnection properties and the knowledge gained the board level reliability. Both are the primary acceptance criteria for implementation. Knowledge about melting characteristic, surface tension depend on temperature and organic vehicles, wetting behavior, electrical conductivity, thermal conductivity, specific heat, mechanical strength, creep and relaxation properties, interactions to preferred finishes (minor impurities), intermetallic growth, content of IMC, brittleness depend on solved elements/IMC, fatigue resistance, damage mechanism, affinity against oxygen, reduction potential, decontamination efforts, endo-/exothermic reactions, diffusion properties related to finishes or bare materials, isothermal fatigue, thermo-cyclic fatigue, corrosion properties, lifetime prediction based on board level results, compatibility with rework/repair solders, rework temperatures of modified solders (Impurities, change in the melting point or range), compatibility to components and laminates.

• Proceedings of the International Microelectronics And Packaging Society Conference
• /
• 2004.09a
• /
• pp.211-219
• /
• 2004

A new alloy definition will be presented concerning increasing demands for the board level reliability of miniaturized interconnections. The damage mechanism for LFBGA components on different board finishes is not quite understood. Further demands from mobile phones are the drop test, characterizing interface performance of different package constructions in relation to decreased pad constructions and therefore interfaces. The paper discusses the characterization of interfaces based on SnPb, SnPbXYZ, SnAgCu and SnAgCuInNd ball materials and SnAgCuInNd as solder paste, the stability after accelerated tests and the description of modified interfaces stric시y related to the assembly conditions, dissolution behavior of finishes on board side and the influence of intermetallic formation. The type of intermetallic as well as the quantity of intermetallics are observed, primaliry the hardness, E modules describing the ability of strain/stress compensation. First results of board level reliability are presented after TCT-40/+150. Improvement steps from the ball formulation will be discussed in conjunction to the implementation of lead free materials. In order to optimize ball materials for area array devices accelareted aging conditions like TCTs were used to analyze the board level reliability of different ball materials for BGA, LFBGA, CSP, Flip Chip. The paper outlines lead-free ball analysis in comparison to conventional solder balls for BGA and chip size packages. The important points of interest are the description of processability related to existing ball attach procedures, requirements of interconnection properties and the knowledge gained the board level reliability. Both are the primary acceptance criteria for implementation. Knowledge about melting characteristic, surface tension depend on temperature and organic vehicles, wetting behavior, electrical conductivity, thermal conductivity, specific heat, mechanical strength, creep and relaxation properties, interactions to preferred finishes (minor impurities), intermetallic growth, content of IMC, brittleness depend on solved elements/IMC, fatigue resistance, damage mechanism, affinity against oxygen, reduction potential, decontamination efforts, endo-/exothermic reactions, diffusion properties related to finishes or bare materials, isothermal fatigue, thermo-cyclic fatigue, corrosion properties, lifetime prediction based on board level results, compatibility with rework/repair solders, rework temperatures of modified solders (Impurities, change in the melting point or range), compatibility to components and laminates.

• Fashion & Textile Research Journal
• /
• v.19 no.2
• /
• pp.230-239
• /
• 2017

In the clothes industry, there are lack of experts including technical designers who can analyze the fit of clothes. This study is to provide practical data available for fit analysis education by distinguishing the differences in standards and aspects of garment fit between experts and novices, through the eye-tracking technology to quantify the sense of fit. For this study, two groups were organized; one composed of 7 experts with over 15 year-experience including technical designers and patternmakers, and the other composed of 7 novices who are students majoring in clothing. Wearing the goggle type eye-tracker Tobii Pro Glasses 2, the participants in the experiments were required to conduct fit analyses for a pair of pants on a live model. After those experiments, they were required to check the items for fit analysis and assess the importance level of them on a questionnaire. The differences between the two groups in the ratios of total visit count and total visit duration by each AIO(Area of Interest) of clothes were analyzed through non-parametric statistical test. The results of eye tracking experiments showed that experts focused on center front and back line, crotch area, and side seam, while novice's fixation points were dispersed around the pants. The survey results showed that the experts put importance on the center line position and its verticality, front-back proportion of side seam line, and front-back proportion of waist line, 71.4~100% of whom checked them, while 14.3% of the novices checked them.