• Journal of Broadcast Engineering
• /
• v.26 no.2
• /
• pp.125-131
• /
• 2021

In recent years, Convolutional Neural Networks (CNNs) have achieved outstanding performance in the fields of computer vision such as image classification, object detection, visual quality enhancement, etc. However, as huge amount of computation and memory are required in CNN models, there is a limitation in the application of CNN to low-power environments such as mobile or IoT devices. Therefore, the need for neural network compression to reduce the model size while keeping the task performance as much as possible has been emerging. In this paper, we propose a method to compress CNN models by combining matrix decomposition methods of LR (Low-Rank) approximation and CP (Canonical Polyadic) decomposition. Unlike conventional methods that apply one matrix decomposition method to CNN models, we selectively apply two decomposition methods depending on the layer types of CNN to enhance the compression performance. To evaluate the performance of the proposed method, we use the models for image classification such as VGG-16, RestNet50 and MobileNetV2 models. The experimental results show that the proposed method gives improved classification performance at the same range of 1.5 to 12.1 times compression ratio than the existing method that applies only the LR approximation.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.18 no.1
• /
• pp.74-82
• /
• 2019

The revolution of industry 4.0 is enabling us to build an intelligent connection society called smart cities. The use of renewable energy in particular solar energy is extremely important for modern society due to the growing power demand in smart cities, but its difficult to monitor and manage in each buildings since need to be deploy low energy sensors and information need to be transfer via wireless sensor network (WSN). The Internet of Things (IoT) / low-power wide-area (LPWA) is an emerging WSN technology, to collect and monitor data about environmental and physical electrical / electronics devices conditions in real time. However, providing power to IoT sensor end devices and other public electrical loads such as street lights, etc is an important challenging role because the sensor are usually battery powered and have a limited life time. In this paper, we proposes an efficient solar energy-based power management scheme for smart city based on IoT technology using LoRa wide-area network (LoRaWAN). This approach facilitates to maintain and prevent errors of solar panel based energy systems. The proposed solution maximizing output the power generated from solar panels system to distribute the power to the load and the grid. In this paper, we proved the efficiency of the proposed system with Simulink based system modeling and real-time emulation.

• Journal of the Korean Society for information Management
• /
• v.36 no.3
• /
• pp.149-174
• /
• 2019

This study analyzes the actual use data of the websites of university libraries, analyzes the users' usage behavior, and proposes improvement measures for the websites. The study analyzed users' traffic and analyzed their usage behavior from January 2018 to December 2018 on the C University website. The website's analysis tool used 'Google Analytics'. The web traffic variables were analyzed in five categories: user general characteristics, user environment analysis, visit analysis, inflow analysis, site analysis, and site analysis based on the metrics of sessions, users, page views, pages per session, average session time, and bounce rate. As a result, 1) In the analysis results of general characteristics of users, there was some access to the website not only in Korea but also in China. 2) In the user experience analysis, the main browser type appeared as Internet Explorer. The next place was Chrome, with a bounce rate of Safari, third and fourth, double that of the Explore or Chrome. In terms of screen resolution, 1920x1080 resolution accounted for the largest percentage, with access in a variety of other environments. 3) Direct inflow was the highest in the inflow media analysis. 4) The site analysis showed the most page views out of 4,534,084 pages, followed by the main page, followed by the lending/extension/history/booking page, the academic DB page, and the collection page.

• Management & Information Systems Review
• /
• v.37 no.4
• /
• pp.41-62
• /
• 2018

As the volume of unstructured data increases through various social media, Internet news articles, and blogs, the importance of text analysis and the studies are increasing. Since text analysis is mostly performed on a specific domain or topic, the importance of constructing and applying a domain-specific dictionary has been increased. The quality of dictionary has a direct impact on the results of the unstructured data analysis and it is much more important since it present a perspective of analysis. In the literature, most studies on text analysis has emphasized the importance of dictionaries to acquire clean and high quality results. However, unfortunately, a rigorous verification of the effects of dictionaries has not been studied, even if it is already known as the most essential factor of text analysis. In this paper, we generate three dictionaries in various ways from 39,800 news articles and analyze and verify the effect each dictionary on the accuracy of document classification by defining the concept of Intrinsic Rate. 1) A batch construction method which is building a dictionary based on the frequency of terms in the entire documents 2) A method of extracting the terms by category and integrating the terms 3) A method of extracting the features according to each category and integrating them. We compared accuracy of three artificial neural network-based document classifiers to evaluate the quality of dictionaries. As a result of the experiment, the accuracy tend to increase when the "Intrinsic Rate" is high and we found the possibility to improve accuracy of document classification by increasing the intrinsic rate of the dictionary.

• The Journal of the Korea Contents Association
• /
• v.21 no.10
• /
• pp.286-308
• /
• 2021

For a long time, dramas that everyone has enjoyed at home have become the most popular cultural contents due to the development of digital technology and the influence of Hallyu.(Korean Wave) This study was conducted in-depth interviews and participatory observations on the background, role, identity, and labor experience of TV planning producers who appeared in the drama production process with the implementation of outsourcing production policy in 1991. The number of dramas produced increased sharply in the mid-2000s due to the Korean Wave. Against this backdrop, the planning producer has expanded their scope in the drama production process and emerged as a new drama producer. The planning producer plays a role in creating an environment in which writers and directors can be selected with the identity of "not a creator but a producer of dramas" and lead drama planning. OTT and watching TV on the Internet have made it possible to watch dramas without TV. As this phenomenon accelerates and becomes commonplace, fewer consumers adhere to the traditional way of watching dramas using TV, and consumers' emotional tastes become more demanding. In this environment, TV planning producers are leading the production of dramas, exerting as much influence as writers and directors. They are also building new power relationships among drama producers by securing planning and financial power.

• Journal of Science Education
• /
• v.45 no.1
• /
• pp.69-89
• /
• 2021

The purpose of this study is to examine personal characteristics of two elementary students who developed individual interest in science. 201 students of two elementary schools in Seoul participated in questionnaire survey engaged in to investigate students' interest in science, engagements and preference in science-related-activities three times a year. This case study was conducted with seven students who showed typical types of interest in science. Students wrote a photo-journal for 12 weeks. They were interviewed every other week by a researcher. We analyzed two students who developed individual interest. It turns out that RF participated in doing experiments at home, and disassembling and assembling things. Participation of activities was a process of solving curiosity. When he was unable to solve problems, he searches the internet or books. Sometimes he would ask for help from his teacher. JW engaged in activities: drawing, doing experiments at home, and going to informal education center. She communicated with others by using online-offline media. Curious questions were solved by herself. The two students have a tendency to solve problems by themselves. Also, positive science-related activities were going on at home and in school. This study not only give insights and understanding of students who developed individual interest in science but also provide implication for educators to examine personal characteristics of students.

• Journal of the Korea Society of Computer and Information
• /
• v.26 no.4
• /
• pp.105-112
• /
• 2021

In the era of big data, interest in data is exploding. In particular, the development of the Internet and social media has led to the creation of new data, enabling the realization of the era of big data and artificial intelligence and opening a new chapter in convergence technology. Also, in the past, there are many demands for analysis of data that could not be handled by programs. In this paper, an analysis model was designed and verified for classification of unstructured data, which is often required in the era of big data. Data crawled DBPia's thesis summary, main words, and sub-keyword, and created a database using KoNLP's data dictionary, and tokenized words through morpheme analysis. In addition, nouns were extracted using KAIST's 9 part-of-speech classification system, TF-IDF values were generated, and an analysis dataset was created by combining training data and Y values. Finally, The adequacy of classification was measured by applying three analysis algorithms(random forest, SVM, decision tree) to the generated analysis dataset. The classification model technique proposed in this paper can be usefully used in various fields such as civil complaint classification analysis and text-related analysis in addition to thesis classification.

• Industry Promotion Research
• /
• v.6 no.1
• /
• pp.63-77
• /
• 2021

This study analyzed the impressional characteristics and commonality of the seven people who reached the final final in the 2020 TV Chosun Mister Trot contest, which is the result of the success of the entertainer through the impression. The analysis criteria were set by referring to the classic Ma-uisangbeob and the academic papers on Impressionism, and the faces of seven subjects were collected from the Internet and media. The results of the analysis showed that the following common points were found: First, hair was developed on both sides of the forehead rather than the development of the forehead, which is suitable for arts and physical education rather than studying. Second, most of them had eyebrow bones [the brains of the brain] and cartilage inside their ears was protruding, which is a type of success through effort, deciding on their own life. Third, the mouth was large, the lips were straight, and the neck was thick. This also showed the temperament and talent of artistic ability. Fourth, it developed greatly in the corresponding part by age. In conclusion, while the inherently innate features of facial features cannot prove all that is fatalistic, the characteristic elements of certain talents were clearly manifested. And what they were able to gain and be loved in the entertainment industry was the result of their own efforts beyond it. These results have implications that can be used to determine their talents or career paths in simple aspects.

• The Journal of the Convergence on Culture Technology
• /
• v.8 no.1
• /
• pp.85-92
• /
• 2022

In order to identify the factors and problems in which military sexual violence is a continuous and repeated blind spot, this study conducted a content analysis focusing on articles of military sexual violence incidents covered in Internet news from January 2010 to June 15, 2021. carried out. As a result of the study, structurally unequal power relations, authoritarian and closed military organizational culture, internal military response system that is distrustful of passive responses to sexual violence, and enveloping family-friendly investigations and tolerant punishment of perpetrators are blind spots despite the Ministry of National Defense's efforts to improve. factors that exist. Underlying this, the compensatory spirit caused by the conscription system and the negative effects of the patriarchal system are embodied in the national sentiment, suggesting that the sense of crisis of division and an overly permissive attitude toward the military act as a factor that slows change. As an improvement plan according to the results, it is necessary to entail the establishment of a civilian-centered judicial institution, strong punishment of perpetrators, and limited pension payment, as well as honorable punishment such as 'class demotion' in the military culture with a clear hierarchical relationship. Taken together, we can see that most military sexual violence is caused by a hierarchy, and it strongly suggests that the main cause of sexual violence is unequal power relations.

• Journal of the Korea Society of Computer and Information
• /
• v.27 no.4
• /
• pp.169-181
• /
• 2022

Currently, due to the activation of SNS live broadcasting, online services based on influencer live broadcasting are becoming a major consumption trend around the world. This study aims to verify the relationship between service quality, customer satisfaction, product characteristics, and acceptance intention for influencer broadcasting based on nfluencer broadcasting experiences in an Internet environment. This study conducted a survey of users who experienced live broadcasting on social media in Taiwan from June 29 to August 30, 2020, and a total of 253 copies were used for empirical analysis. The collected data were analyzed through SPSS 25.0. The results of the empirical analysis are summarized as follows. First, it was found that the service quality factors (reliability, tangibility, responsiveness, certainty, and empathy) of Taiwan's influencer live broadcast had a significant effect on live broadcast satisfaction. Second, it was found that the product characteristics of Taiwan's influencer live broadcasting had a significant effect on product satisfaction. Third, it was found that live broadcasting satisfaction and product satisfaction had a significant effect on the acceptance intention of new brands in Taiwan's influencer live broadcasting. This study will provide useful data for establishing efficient marketing strategies to improve live and product satisfaction and increase acceptance of new brands by identifying service quality factors and product characteristics of Taiwan's influencer Live Broadcasting.