• Journal of Intelligence and Information Systems
• /
• v.24 no.3
• /
• pp.1-19
• /
• 2018

Large amount of data is now available for research and business sectors to extract knowledge from it. This data can be in the form of unstructured data such as audio, text, and image data and can be analyzed by deep learning methodology. Deep learning is now widely used for various estimation, classification, and prediction problems. Especially, fashion business adopts deep learning techniques for apparel recognition, apparel search and retrieval engine, and automatic product recommendation. The core model of these applications is the image classification using Convolutional Neural Networks (CNN). CNN is made up of neurons which learn parameters such as weights while inputs come through and reach outputs. CNN has layer structure which is best suited for image classification as it is comprised of convolutional layer for generating feature maps, pooling layer for reducing the dimensionality of feature maps, and fully-connected layer for classifying the extracted features. However, most of the classification models have been trained using online product image, which is taken under controlled situation such as apparel image itself or professional model wearing apparel. This image may not be an effective way to train the classification model considering the situation when one might want to classify street fashion image or walking image, which is taken in uncontrolled situation and involves people's movement and unexpected pose. Therefore, we propose to train the model with runway apparel image dataset which captures mobility. This will allow the classification model to be trained with far more variable data and enhance the adaptation with diverse query image. To achieve both convergence and generalization of the model, we apply Transfer Learning on our training network. As Transfer Learning in CNN is composed of pre-training and fine-tuning stages, we divide the training step into two. First, we pre-train our architecture with large-scale dataset, ImageNet dataset, which consists of 1.2 million images with 1000 categories including animals, plants, activities, materials, instrumentations, scenes, and foods. We use GoogLeNet for our main architecture as it has achieved great accuracy with efficiency in ImageNet Large Scale Visual Recognition Challenge (ILSVRC). Second, we fine-tune the network with our own runway image dataset. For the runway image dataset, we could not find any previously and publicly made dataset, so we collect the dataset from Google Image Search attaining 2426 images of 32 major fashion brands including Anna Molinari, Balenciaga, Balmain, Brioni, Burberry, Celine, Chanel, Chloe, Christian Dior, Cividini, Dolce and Gabbana, Emilio Pucci, Ermenegildo, Fendi, Giuliana Teso, Gucci, Issey Miyake, Kenzo, Leonard, Louis Vuitton, Marc Jacobs, Marni, Max Mara, Missoni, Moschino, Ralph Lauren, Roberto Cavalli, Sonia Rykiel, Stella McCartney, Valentino, Versace, and Yve Saint Laurent. We perform 10-folded experiments to consider the random generation of training data, and our proposed model has achieved accuracy of 67.2% on final test. Our research suggests several advantages over previous related studies as to our best knowledge, there haven't been any previous studies which trained the network for apparel image classification based on runway image dataset. We suggest the idea of training model with image capturing all the possible postures, which is denoted as mobility, by using our own runway apparel image dataset. Moreover, by applying Transfer Learning and using checkpoint and parameters provided by Tensorflow Slim, we could save time spent on training the classification model as taking 6 minutes per experiment to train the classifier. This model can be used in many business applications where the query image can be runway image, product image, or street fashion image. To be specific, runway query image can be used for mobile application service during fashion week to facilitate brand search, street style query image can be classified during fashion editorial task to classify and label the brand or style, and website query image can be processed by e-commerce multi-complex service providing item information or recommending similar item.

• Journal of Intelligence and Information Systems
• /
• v.19 no.2
• /
• pp.101-123
• /
• 2013

Recently due to the advancements of science and information technology, the socio-economic business areas are changing from the industrial economy to a knowledge economy. Furthermore, companies need to do creation of new value through continuous innovation, development of core competencies and technologies, and technological convergence. Therefore, the identification of major trends in technology research and the interdisciplinary knowledge-based prediction of integrated technologies and promising techniques are required for firms to gain and sustain competitive advantage and future growth engines. The aim of this paper is to understand the recent research trend in management of technology (MOT) and to foresee promising technologies with deep knowledge for both technology and business. Furthermore, this study intends to give a clear way to find new technical value for constant innovation and to capture core technology and technology convergence. Bibliometrics is a metrical analysis to understand literature's characteristics. Traditional bibliometrics has its limitation not to understand relationship between trend in technology management and technology itself, since it focuses on quantitative indices such as quotation frequency. To overcome this issue, the network focused bibliometrics has been used instead of traditional one. The network focused bibliometrics mainly uses "Co-citation" and "Co-word" analysis. In this study, a keywords network analysis, one of social network analysis, is performed to analyze recent research trend in MOT. For the analysis, we collected keywords from research papers published in international journals related MOT between 2002 and 2011, constructed a keyword network, and then conducted the keywords network analysis. Over the past 40 years, the studies in social network have attempted to understand the social interactions through the network structure represented by connection patterns. In other words, social network analysis has been used to explain the structures and behaviors of various social formations such as teams, organizations, and industries. In general, the social network analysis uses data as a form of matrix. In our context, the matrix depicts the relations between rows as papers and columns as keywords, where the relations are represented as binary. Even though there are no direct relations between papers who have been published, the relations between papers can be derived artificially as in the paper-keyword matrix, in which each cell has 1 for including or 0 for not including. For example, a keywords network can be configured in a way to connect the papers which have included one or more same keywords. After constructing a keywords network, we analyzed frequency of keywords, structural characteristics of keywords network, preferential attachment and growth of new keywords, component, and centrality. The results of this study are as follows. First, a paper has 4.574 keywords on the average. 90% of keywords were used three or less times for past 10 years and about 75% of keywords appeared only one time. Second, the keyword network in MOT is a small world network and a scale free network in which a small number of keywords have a tendency to become a monopoly. Third, the gap between the rich (with more edges) and the poor (with fewer edges) in the network is getting bigger as time goes on. Fourth, most of newly entering keywords become poor nodes within about 2~3 years. Finally, keywords with high degree centrality, betweenness centrality, and closeness centrality are "Innovation," "R&D," "Patent," "Forecast," "Technology transfer," "Technology," and "SME". The results of analysis will help researchers identify major trends in MOT research and then seek a new research topic. We hope that the result of the analysis will help researchers of MOT identify major trends in technology research, and utilize as useful reference information when they seek consilience with other fields of study and select a new research topic.

• KDI Journal of Economic Policy
• /
• v.13 no.2
• /
• pp.97-127
• /
• 1991

This paper investigates the roles and the limitations of the Korean antitrust agencies-the Office of Fair Trade (OFT) and the Fair Trade Commission (FTC) during the making of the major industrial policies of the 1980's. The Korean antitrust agencies played only a minimal role in three major industrial policy-making issues in the 1980's- the enactment of the Industrial Development Act (IDA), the Industrial Rationalization Measures according to the IDA, and the Industrial Readjustment Measures on Consolidation of Large Insolvent Enterprises based on the revised Tax Exemption and Reduction Control Act. As causes for this performance bias in the Korean antitrust system, this paper considers five factors according to the current literature on implementation failure: ambiguous and insufficient statutory provisions of the Monopoly Regulation and Fair Trade Act (MRFTA); lack of resources; biased attitudes and motivations of the staff of the OFT and the FTC; bureaucratic incapability; and widespread misunderstanding about the roles and functions of the antitrust system in Korea. Among these five factors, bureaucratic incompetence and lack of understanding in various policy implementation environments about the roles and functions of the antitrust system have been regarded as the most important ones. Most staff members did not have enough educational training during their school years to engage in antitrust and fair trade policy-making. Furthermore, the high rate of staff turnover due to a mandatory personnel transfer system has prohibited the accumulation of knowledge and skills required for pursuing complicated structural antitrust enforcement. The limited capability of the OFT has put the agency in a disadvantaged position in negotiating with other economic ministries. The OFT has not provided plausible counter-arguments based on sound economic theories against other economic ministries' intensive market interventions in the name of rationalization and readjustment of industries. If the staff members of antitrust agencies have lacked substantive understanding of the antitrust and fair trade policy, the rest of government agencies must have had serious problems in understanding the correst roles and functions of the antitrust system. The policy environment of the Korean antitrust system, including other economic ministries, the Deputy Prime Minister, and President Chun, have tended to conceptualize the OFT more as an agency aiming only at fair trade policy and less as an agency that should enforce structural monopoly regulation as well. Based on this assessment of the performance of the Korean antitrust system, this paper evaluate current reform proposals for the MRFT A. The inclusion of the regulation of conglomerate mergers and of business divestiture orders may be a desirable revision, giving the MRFTA more complete provisions. However, given deficient staff experties and the unfavorable policy environments, it would be too optimistic and naive to expect that the inclusion of these provisions alone could improve the performance of the Korean antitrust system. In its conclusion, this paper suggests several policy recommendations for the Korean antitrust system, which would secure the stable development and accumulation of antitrust expertise for its staff members and enough understanding and conformity from its environments about its antitrust goals and functions.

• 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 the Korean Institute of Gas
• /
• v.14 no.3
• /
• pp.26-34
• /
• 2010

KOGAS, established in 1983 by law to ensure stable gas supply to the public, is responsible for the wholesale distribution to 30 city gas companies that deal with the retail distribution of natural gas in their geographic areas. The gas imported by KOGAS is measured by checking the level difference of LNG shipped in tankers before and after unloading. The analysis of gas composition is essential because the imported gas price is determined by its calorific value. The turbine meter is widely used for measuring the gas sold to city gas companies. Unlike the metering system for power plants, there is no gas chromatograph since the custody transfer of gas to the city gas companies is not billed by calorific value, but by volume basis. The gas quantity that a city gas company has bought from KOGAS is not equal to the quantity that the company sold to its customers. There have been some discrepancies between the wholesale gas meter readouts and retail ones due to some inherent errors of meters and some operational issues of the meters. This paper investigates the controversies regarding the real quantity of gas between distributors and consumers. It will discus and suggest desirable policies, both technically and economically, in order to solve the discrepancies of gas measurement.

• Journal of the Korean Society for Library and Information Science
• /
• v.50 no.3
• /
• pp.309-327
• /
• 2016

Scholarly monographs have been a main vehicle for knowledge representation and transfer as well as an important research outcome. University presses have long been considered as the last bastion for scholarly monographs that have low commercial prospect. Until now there has not been a systematic data analysis regarding scholarly monograph production by university presses in Korea. In this paper, we collected bibliographic records of university presses' monograph publications between 1950 and 2015 using the National Library of Korea's online catalog system. A total of 21,015 records were used in the analysis. In particular, for monographs published between 2001 and 2015, we categorized them into scholarly monographs and non-scholarly monographs. University presses' publishing showed sharp increase during the 1990's but is in decline after its peak in 2005. University presses seem to have engaged in publishing more non-scholarly monographs than scholarly monographs by a ratio of 6:4. Large university presses in size seem to produce higher proportions of scholarly monographs than smaller presses. In terms of authoring types, single authorship accounts for the highest proportion and on the increase. However, edited books are losing ground as translated books seem to hold steady. Monographs in social sciences have been published more frequently than any other subject areas as there seem to be significant discrepancies among subject areas in terms of the scholarly monograph proportion.

• Korean Chemical Engineering Research
• /
• v.49 no.1
• /
• pp.41-46
• /
• 2011

We investigated experimentally the properties of dust explosion through lycopodium particle clouds in a duct to provide the fundamental knowledge. Propagating dust flames in the vertical duct of 120 cm height and 12 cm square cross-section were observed by digital video camera and flame front is visualized using by PIV(Particle Image Velocimetry) system. As the result, when the same average dust concentration existed in the vertical duct, downward flame propagation was faster than the upward flame propagation, its rate increased with dust concentration in 300g/$m^3$. Post flames were caused by the ignition of unburned particles which flowed into the rear region of flame from passage between flame and duct wall, and they generated regardless of duct condition. Also, it was found that appearance frequency of post flames during dust flame propagations increased with the increase of dust concentration.

• Journal of the Korean Society for Railway
• /
• v.19 no.1
• /
• pp.20-28
• /
• 2016

Curved squeal noise may result when railway vehicles run on curved tracks. Contact between the wheels and the rails causes a stick-slip phenomenon, which generates squeal noise. In order to identify the mechanism of the squeal noise systematically, a scaled test rig has been fabricated. Knowledge of the contact forces between the wheels and the rail rollers is essential for investigating the squeal noise characteristics; however, it is difficult to measure there contact force. In this study, contact forces have been calculated indirectly according to the modal behavior of the subframe that supports the rail roller and the responses at specific positions of that subframe. In order to verify the estimated contact forces, the displacements at the contact points between the wheels and rail rollers have been calculated from the estimated forces; the resulting values have been compared with the measured displacement values. The SPL at the specific location has been calculated using the estimated contact forces and this also has been compared with the SPL, measured in a semi-anechoic chamber. The comparisons in displacements and SPLs show good correlation.

• Korean Journal of Social Welfare
• /
• v.52
• /
• pp.89-115
• /
• 2003

While many social work studies has focused on the development of practice intervention and demonstration of the efficacy, little attention has been given to the way how the results can be disseminated and utilized in practice area. But attempts to bridge the research-practice gap empathize importance of dissemination and utilization including knowledge utilization, innovation diffusion, information dissemination, program replication, and technology transfer etc. In this recognition, this study that focused on the process after development of program tried to search and analyze the real disseminating process and utilization state with applying Growing-up Skills Program(GUSP) developed and disseminated by this researcher. For the purpose, theories and models for program diffusion were reviewed, and dissemination & utilization of GUSP was analyzed on the 'Herie & Martin's model' with retrospective perspectives. Through this tracer study, these were confirmed or founded as problems of GUSP in disseminating process : lack of specific strategy for the dissemination ; unplanned coping due to little preparation ; and unclear and insufficient understanding of the target system in the beginning stage of dissemination. And followings are suggested for the improvement of social work program dissemination & utilization in the field: to conduct diffusion process on the basis of a proper model; to consider integrative relationship between D&D and dissemination & utilization ; to endeavor for collecting feed back from the field ; and to reinforce social work education and study related to innovation diffusion. These results showed importance of dissemination & utilization in social work and utility of GUSP despite of several limitations. More concerns of dissemination & utilization are needed for the integration of research and practice in Korean social work.

• Korean Journal of Acupuncture
• /
• v.40 no.3
• /
• pp.79-89
• /
• 2023

Objectives : The purpose of this study is to investigate various application methods of focused ultrasound and apply them to clinical use in Korean medicine. Methods : Search was performed using the search engines of electronic databases, including PubMed, ScienceDirect, Cumulative Index to Nursing and Allied Health Literature (CINAHL), ScienceON, Oriental Medicine Advanced Searching Integrated System (OASIS), China National Knowledge Infrastructure (CNKI), Wanfang Data, Japan Science Technology Information Aggregator, Electronic (J-STAGE) and Citation Information by NII (CiNii), from inception to July 2023 without language limitation. Inclusion criteria were clinical studies including randomized controlled trials (RCTs), and animal experimental studies related with focused ultrasound treatments for acupoints or meridian sinews. Results : Total 17 papers, 7 for RCT, 6 for in vivo animal studies, and other experimental studies, were finally selected. Indications used in studies were shoulder pain, back pain, chronic back pain, and degenerative knee arthritis. In experimental studies, studies on animal models of hypoxic ischemic brain damage and hyperlipidemia were also conducted. As for the acupoints, LR3, LI4, and ST36 were used in clinical studies and, in animal experimental studies, GV20, KI1, and ST36 were used. As for the dose, 4 studies below 3 W/cm² and 3 studies in the range of 0.625 to 5 W/cm² in clinical studies, and all studies did not exceed 5 W/cm². In animal experimental studies, 0.5 W/cm², 2 W/cm², 7.5 WW/cm², 15 W/cm², 10~20 W/cm² were used. In all three studies describing the penetration depth during irradiation, it was less than 1 cm. Conclusions : We suggest that focused ultrasound is an appropriate treatment tool for stimulating the acupoints to transfer heat energy. Future studies with rigorous and well-designed RCTs for various diseases will be required to ascertain the focused ultrasound stimulate acupoints or meridian sinews.