• Journal of KIISE:Software and Applications
• /
• v.36 no.4
• /
• pp.253-260
• /
• 2009

Many algorithms for computer vision and pattern recognition have recently been implemented on GPU (graphic processing unit) for faster computational times. However, the implementation has two problems. First, the programmer should master the fundamentals of the graphics shading languages that require the prior knowledge on computer graphics. Second, in a job that needs much cooperation between CPU and GPU, which is usual in image processing and pattern recognition contrary to the graphic area, CPU should generate raw feature data for GPU processing as much as possible to effectively utilize GPU performance. This paper proposes more quick and efficient implementation of neural networks on both GPU and multi-core CPU. We use CUDA (compute unified device architecture) that can be easily programmed due to its simple C language-like style instead of GPU to solve the first problem. Moreover, OpenMP (Open Multi-Processing) is used to concurrently process multiple data with single instruction on multi-core CPU, which results in effectively utilizing the memories of GPU. In the experiments, we implemented neural networks-based text extraction system using the proposed architecture, and the computational times showed about 15 times faster than implementation on only GPU without OpenMP.

• The Journal of Society for e-Business Studies
• /
• v.22 no.1
• /
• pp.123-146
• /
• 2017

The information that consumers share in the SNS (Social Networking Service) has a great influence on the purchase of consumers. Therefore, it is necessary to pay attention to new research methodology and advertising strategy using Social Big Data. In this context, the purpose of this study is to quantitatively analyze customer value through Social Big Data. In this study, we analyzed the value structure of consumers for the three smartphone brands through text mining and positive/negative image analysis. Analysis result, it was possible to distinguish the emotional aspects (sensitivity) and rational aspects (rationality) for customer value per brand. In the case of the Galaxy S7 and iPhone 6S, emotional aspects were important before the launch, but the rational aspects was important after release date. On the other hand, in the case of the LG G5, emotional aspects were important before and after launch. We can propose two core advertising strategies based on analyzed consumer value. When developing advertising strategy in the case of the Galaxy S7, there is a need to emphasize the rational aspects of product attributes and differentiated functions. In the case of the LG G5, it is necessary to consider the emotional aspects of happiness, excitement, pleasure, and fun that are felt by using products in advertising strategy. As a result, this study will provide a good standard for actual advertising strategy through consumer value analysis. Advertising strategies are primarily driven by intuition or experience. Therefore, it is important to develop advertising strategies by analyzing consumer value through social big data analysis.

• Journal of the Korea Convergence Society
• /
• v.9 no.1
• /
• pp.419-431
• /
• 2018

In this paper, we propose a novel method for monitoring mood trend of Twitter users by analyzing their daily tweets for a long period. Then, to more accurately understand their tweets, we analyze all types of content in tweets, i.e., texts and emoticons, and images, thus develop a multimodal sentiment analysis method. In the proposed method, two single-modal analyses first are performed to extract the users' moods hidden in texts and images: a lexicon-based and learning-based text classifier and a learning-based image classifier. Thereafter, the extracted moods from the respective analyses are combined into a tweet mood and aggregated a daily mood. As a result, the proposed method generates a user daily mood flow graph, which allows us for monitoring the mood trend of users more intuitively. For evaluation, we perform two sets of experiment. First, we collect the data sets of 40,447 data. We evaluate our method via comparing the state-of-the-art techniques. In our experiments, we demonstrate that the proposed multimodal analysis method outperforms other baselines and our own methods using text-based tweets or images only. Furthermore, to evaluate the potential of the proposed method in monitoring users' mood trend, we tested the proposed method with 40 depressive users and 40 normal users. It proves that the proposed method can be effectively used in finding depressed users.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2008.10a
• /
• pp.524-527
• /
• 2008

Recently, digital and the Internet are widespread out of the world, and multimedia processing technology and the development of information and communication technology in education using the Internet as the demand is rapidly increasing. Also, we tan easily use informations with less restrictions of time and space. however, several kinds of audio, media to integrate multimedia data, such as the proliferation of demands for representation. Therefore, in 1998, W3C presented an international standard, SMIL in order to solve multimedia object representation and synchronization problems. By using SMIL, various multimedia elements can be integrated as a multimedia document with proper view in a spate and time. Using this SMIL document, we can create new internet radio broadcasting service that delivers not noly audio data but also various text, image and video. In this paper, with the system, teachers can easily create multimedia courseware and living broadcast their torture on network, students can receive audio-video information of the teacher, screen displays of the teachers computer. Moreover students can communicate with teacher simultaneously by text editor windows. Students can also order courseware after class.

• Journal of the Korean Institute of Traditional Landscape Architecture
• /
• v.36 no.1
• /
• pp.78-88
• /
• 2018

The purpose of this study is to establish a utilization plan according to the cultural value of Placenta Chamber of King Sejong's Princes. We used SNS to analyze various public perceptions and opinions, collected data and analyzed it. The collection period is from June 01, 2007 to June 30, 2017 (for about 10 years), We gathered data from blogs, cafes, and Knowledge IN that contain keywords related to 'Placenta Chamber', 'Placenta Chamber of Seongju', 'Placenta Chamber of King Sejong's Princes'. and Analyzed using the text mining method of the big date program. Based on the main results of the big data analysis, Placenta Chamber's method of utilization was derived. As a result, major keywords such as King Sejong Great, Prince, Sungju, Feng Shui, culture, preservation, blessing etc were derived. The association of 'world', 'heritage', 'cultural heritage' is high, and the connection of 'Placenta Chamber', 'Gyeongsangbuk-do', 'cultural property' is high, and it was able to confirm the value of Placenta Chamber as a world cultural heritage. and It is necessary to induce visitors to feel stimulation or change of surroundings through facility refurbishment and environmental improvement around Placenta Chamber.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.7 no.6
• /
• pp.1047-1055
• /
• 2006

We patterned nano-width lines on a super hard bulk diamond substrate by varying the ion beam current and ion beam sources with a dual beam field ion beam (FIB). In addition, we successfully fabricated two-dimensional nano patterns and three-dimensional nano plate modules. We prepared nano lines on a diamond and a silicon substrate at the beam condition of 30 kV, 10 pA $\sim$ 5 nA with $Ga^+$ ion and $H_2O$ assisted ion sources. We measured each of the line-width, line-depth, etched line profiles, etch rate, and aspect ratio, and then compared them. We confirmed that nano patterning was possible on both a bulk diamond and a silicon substrate. The etch rate of $H_2O$ source can be enhanced about two times than that of Ga source. The width of patterns on a diamond was smaller than that on a silicon substrate at the same ion beam power The sub-100 nm patterns on a diamond were made under the charge neutralization mode to prevent charge accumulation. We successfully made a two-dimensional, 240 nm-width text of the 300-lettered Lord's Prayer on a gem diamond with 30 kV-30 pA FIB. The patterned text image was readable with a scanning electron microscope. Moreover, three dimensional nano-thick plate module fabrication was made successfully with an FIB and a platinum deposition, and electron energy loss spectrum (EELS) analysis was easily performed with the prepared nano plate module.

• Journal of Broadcast Engineering
• /
• v.27 no.3
• /
• pp.391-401
• /
• 2022

As deep learning-based object detection and recognition research have been developed recently, the scope of application to industry and real life is expanding. But deep learning-based systems in the construction system are still much less studied. Calculating materials in the construction system is still manual, so it is a reality that transactions of wrong volumn calculation are generated due to a lot of time required and difficulty in accurate accumulation. A fast and accurate automatic drawing recognition system is required to solve this problem. Therefore, we propose an AI-based automatic drawing recognition accumulation system that detects and recognizes steel materials in construction drawings. To accurately detect steel materials in construction drawings, we propose data augmentation techniques and spatial attention modules for improving small object detection performance based on YOLOv4. The detected steel material area is recognized by text, and the number of steel materials is integrated based on the predicted characters. Experimental results show that the proposed method increases the accuracy and precision by 1.8% and 16%, respectively, compared with the conventional YOLOv4. As for the proposed method, Precision performance was 0.938. The recall was 1. Average Precision AP_0.5 was 99.4% and AP_0.5:0.95 was 67%. Accuracy for character recognition obtained 99.9.% by configuring and learning a suitable dataset that contains fonts used in construction drawings compared to the 75.6% using the existing dataset. The average time required per image was 0.013 seconds in the detection, 0.65 seconds in character recognition, and 0.16 seconds in the accumulation, resulting in 0.84 seconds.

• Korean small business review
• /
• v.43 no.4
• /
• pp.27-57
• /
• 2021

How does authenticity influence the process that conducts a company's CSR Strategy? Authenticity, an internal/external alignment condition that an employee feels in relation to an organization, means the decision on how true and beneficial to employees through their experiences, such as thoughts and emotions. Also, it can be understood as a process of meaning formation between the organization's strategy to conduct CSR and the perception of employees conducting CSR. To prove the relation between authenticity and CSR clearly, we used various techniques like Text Mining, Topic Modeling and Semantic network analysis about O corporation's 657 review data, from 2015 to 2021. As a result of the analysis, we find out the special issues and types. The analysis shows that the issue concerning the 'external image' is the biggest characteristic of authenticity perception in other conditions. Furthermore, the types of authenticity perception evaluations are largely divided into acceptance and rejection, in detail, five categories. This study indicates that organizations should consider both external and internal conditions when establishing CSR strategies. In addition, it is necessary to be an interactive circular relationship between the organization and employee, collecting and reflecting employee's perceptions. Finally, this study proposes ways to overcome problems related to interaction.

• KIPS Transactions on Software and Data Engineering
• /
• v.12 no.4
• /
• pp.189-198
• /
• 2023

As crimes frequently occur on the street, the spread of CCTV is increasing. However, due to the shortcomings of passively operated CCTV, the need for intelligent CCTV is attracting attention. Due to the heavy system of such intelligent CCTV, high-performance devices are required, which has a problem in that it is expensive to replace the general CCTV. To solve this problem, an intelligent CCTV system that recognizes low-quality images and operates even on devices with low performance is required. Therefore, this paper proposes a Saying CCTV system that can detect threats in real time by using the AWS cloud platform to lighten the system and convert images into text. Based on the data extracted using YOLO v4 and OpenPose, it is implemented to determine the risk object, threat behavior, and threat situation, and calculate the risk using machine learning. Through this, the system can be operated anytime and anywhere as long as the network is connected, and the system can be used even with devices with minimal performance for video shooting and image upload. Furthermore, it is possible to quickly prevent crime by automating meaningful statistics on crime by analyzing the video and using the data stored as text.

• Journal of The Korean Association For Science Education
• /
• v.43 no.2
• /
• pp.73-86
• /
• 2023

This study compared the characteristics of scientific emoticons made by general and science-gifted elementary students. To do this, fifth graders (n=53) at a general elementary school in Gyeonggi province and fifth graders (n=35) at a gifted science education institute in Seoul were selected. Scientific emoticons written by the students were compared according to the number and types. Analysis of the results reveal that the science-gifted students made more scientific emoticons than the general students for thirty minutes. In the comparison of the types of scientific emoticons, there were some similarities and significant differences between general students and science-gifted students. Overall, however, it was found that science-gifted students made more various types of scientific emoticons than general students in 'form' aspects (e.g., generative form of text, descriptive form of text, and expressive form of image) and 'information' aspects (e.g., emotion, construction level, excess of curriculum level, scientific disciplines, and use of scientific knowledge) of the texts and the images in the scientific emoticons. The scientific emoticons made by general and science-gifted elementary students included very few misconceptions. Educational implications of these findings are discussed.