• Asian-Australasian Journal of Animal Sciences
• /
• v.29 no.1
• /
• pp.149-156
• /
• 2016

In this experiment, we proposed and implemented a disease forecasting system using a received signal strength indication ZigBee-based wireless network with a 3-axis acceleration sensor to detect illness at an early stage by monitoring movement of experimentally infected weaned piglets. Twenty seven piglets were divided into control, Salmonella enteritidis (SE) infection, and Escherichia coli (EC) infection group, and their movements were monitored for five days using wireless sensor nodes on their backs. Data generated showed the 3-axis movement of piglets (X-axis: left and right direction, Y-axis: anteroposterior direction, and Z-axis: up and down direction) at five different time periods. Piglets in both infected groups had lower weight gain and feed intake, as well as higher feed conversion ratios than the control group (p<0.05). Infection with SE and EC resulted in reduced body temperature of the piglets at day 2, 4, and 5 (p<0.05). The early morning X-axis movement did not differ between groups; however, the Y-axis movement was higher in the EC group (day 1 and 2), and the Z-axis movement was higher in the EC (day 1) and SE group (day 4) during different experimental periods (p<0.05). The morning X and Y-axis movement did not differ between treatment groups. However, the Z-axis movement was higher in both infected groups at day 1 and lower at day 4 compared to the control (p<0.05). The midday X-axis movement was significantly lower in both infected groups (day 4 and 5) compared to the control (p<0.05), whereas the Y-axis movement did not differ. The Z-axis movement was highest in the SE group at day 1 and 2 and lower at day 4 and 5 (p<0.05). Evening X-axis movement was highest in the control group throughout the experimental period. During day 1 and 2, the Z-axis movement was higher in both of the infected groups; whereas it was lower in the SE group during day 3 and 4 (p<0.05). During day 1 and 2, the night X-axis movement was lower and the Z-axis movement was higher in the infected piglets (p<0.05). Overall, the movement of infected piglets was altered, and the acceleration sensor could be successfully employed for monitoring pig activity.

• Journal of the Korean association of regional geographers
• /
• v.23 no.2
• /
• pp.376-387
• /
• 2017

This paper attempts to analyze the geographical characters of Twitter data and presents analysis potentials for social network analysis in geography. First, this paper suggests a methodology for a topic modeling-based approach in order to identify the geographical characteristics of tweets, including an analysis flow of Twitter data sets, tweet data collection and conversion, textural pre-processing and structural analysis, topic discovery, and interpretation of tweets' topics. GPS coordinates referencing tweets(geotweets) were extracted among sampled Twitter data sets because it contains the tweet place where it was created. This paper identifies a correlated relationship between some specific topics and local places in Jeju. This correlation is closely associated with some place names and local sites in Jeju Island. We assume it is the intention of tweeters to record their tweet places and to share and retweet with other tweeters in some cases. A surface density map shows the hotspots of tweets, detecting around some specific places and sites such as Jeju airport, sightseeing sites, and local places in Jeju Island. The hotspots show similar patterns of the floating population of Jeju, especially the thirty-year age group. In addition, a topic modeling algorithm is applied for the geographical topic discovery and comparison of the spatial patterns of tweets. Finally, this empirical analysis presents that Twitter data, as social network data, provide geographical significance, with topic modeling approach being useful in analyzing the textural features reflecting the geographical characteristics in large data sets of tweets.

• Journal of IKEEE
• /
• v.21 no.4
• /
• pp.412-415
• /
• 2017

In this paper, we propose a multi license plate recognition system using high resolution $360^{\circ}$ omnidirectional IP camera. The proposed system consists of a planar division part of $360^{\circ}$ circular image and a multi license plate recognition part. The planar division part of the $360^{\circ}$ circular image are divided into a planar image with enhanced image quality through processes such as circular image acquisition, circular image segmentation, conversion to plane image, pixel correction using color interpolation, color correction and edge correction in a high resolution $360^{\circ}$ omnidirectional IP Camera. Multi license plate recognition part is through the multi-plate extraction candidate region, a multi-plate candidate area normalized and restore, multiple license plate number, character recognition using a neural network in the process of recognizing a multi-planar imaging plates. In order to evaluate the multi license plate recognition system using the proposed high resolution $360^{\circ}$ omnidirectional IP camera, we experimented with a specialist in the operation of intelligent parking control system, and 97.8% of high plate recognition rate was confirmed.

• Journal of the Institute of Convergence Signal Processing
• /
• v.17 no.2
• /
• pp.83-92
• /
• 2016

In this paper, we introduce the results of the testing and observation systems for performance wideband expansion in the Korean VLBI Network(KVN). The KVN performs VLBI observations to 1024 Mbps data rate, and 8192 Mbps observing for four simultaneous observation is now evaluating for normal operation. The VLBI stations in several world countries developed their own wideband observing systems to observe the celestial objects with high precision and high resolution or are working with several countries. The KVN is planning to introduce a high-speed sampler, OCTAD, for sampling directly up to 2048 MHz bandwidth for RF signal of K/Q/W/D band in the frequency band without conversion. Therefore, as a preliminary study for the performance scalability of the KVN then through the close cooperation with National Astronomical Observatory of Japan (NAOJ), the OCTAD high-speed sampler and OCTADISK2 high-speed recorder were installed in the KVN Yonsei station, and verify the performance through a wideband.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.12 no.5
• /
• pp.467-474
• /
• 2019

In this paper, we propose a novel method to improve FPS while maintaining the accuracy of YOLO v2 model in NVIDIA Jetson TX1. In general, in order to reduce the amount of computation, a conversion to an integer operation or reducing the depth of a network have been used. However, the accuracy of recognition can be deteriorated. So, we use methods to reduce computation and memory consumption through adjustment of the filter size and integrated computation of the network The first method is to replace the $3{\times}3$ filter with a $1{\times}1$ filter, which reduces the number of parameters to one-ninth. The second method is to reduce the amount of computation through CBR (Convolution-Add Bias-Relu) among the inference acceleration functions of TensorRT, and the last method is to reduce memory consumption by integrating repeated layers using TensorRT. For the simulation results, although the accuracy is decreased by 1% compared to the existing YOLO v2 model, the FPS has been improved from the existing 3.9 FPS to 11 FPS.

• Journal of the Korean Society of Environmental Restoration Technology
• /
• v.25 no.1
• /
• pp.25-38
• /
• 2022

Urban problems are constantly occurring around the world due to rapid industrialization and population decline. In particular, as the number of vacant houses is gradually increasing as the population decreases, it is necessary to prepare countermeasures. A plan to utilize vacant houses has emerged to restore the natural environment of the urban ecosystem where forest destruction, damage to habitats of wild animals and plants, and disconnection have occurred due to large-scale development. Through connectivity analysis, it is possible to understand the overall ecosystem flow based on the movement of species and predict the effect when vacant houses are converted into green spaces. Therefore, this study analyzed the green area network to confirm the possibility of greening of vacant houses neglected in Jeonju based on circuit theory. Using Circuitscape and Least-cost path, we tried to identify the connectivity of green areas and propose an ecological axis based on the analysis. In order to apply the resistance values required for analysis based on previous studies, the 2020 subdivision land cover data were integrated into the major classification evaluation items. When the eight forests in the target site were analyzed as the standard, the overall connectivity and connectivity between forests in the area were high, so it is judged that the existing green areas can perform various functions, such as species movement and provision of habitats. Based on the results of the connectivity analysis, the importance of vacant houses was calculated and the top 20 vacant houses were identified, and it was confirmed that the higher the ranking, the more positive the degree of landscape connectivity was when converted to green areas. In addition, it was confirmed that the results of analyzing the least-cost path based on the resistance values such as connectivity analysis and the existing conceptual map showed some differences when comparing the ecological axes in the form. As a result of checking the vacant houses corresponding to the relevant axis based on the width standards of the main and sub-green areas, a total of 30 vacant houses were included in the 200m width and 6 vacant houses in the 80m width. It is judged that the conversion of vacant houses to green space can contribute to biodiversity conservation as well as connectivity between habitats of species as it is coupled with improved green space connectivity. In addition, it is expected to help solve the problem of vacant houses in the future by showing the possibility of using vacant houses.

• Journal of Intelligence and Information Systems
• /
• v.24 no.2
• /
• pp.221-241
• /
• 2018

Deep learning is getting attention recently. The deep learning technique which had been applied in competitions of the International Conference on Image Recognition Technology(ILSVR) and AlphaGo is Convolution Neural Network(CNN). CNN is characterized in that the input image is divided into small sections to recognize the partial features and combine them to recognize as a whole. Deep learning technologies are expected to bring a lot of changes in our lives, but until now, its applications have been limited to image recognition and natural language processing. The use of deep learning techniques for business problems is still an early research stage. If their performance is proved, they can be applied to traditional business problems such as future marketing response prediction, fraud transaction detection, bankruptcy prediction, and so on. So, it is a very meaningful experiment to diagnose the possibility of solving business problems using deep learning technologies based on the case of online shopping companies which have big data, are relatively easy to identify customer behavior and has high utilization values. Especially, in online shopping companies, the competition environment is rapidly changing and becoming more intense. Therefore, analysis of customer behavior for maximizing profit is becoming more and more important for online shopping companies. In this study, we propose 'CNN model of Heterogeneous Information Integration' using CNN as a way to improve the predictive power of customer behavior in online shopping enterprises. In order to propose a model that optimizes the performance, which is a model that learns from the convolution neural network of the multi-layer perceptron structure by combining structured and unstructured information, this model uses 'heterogeneous information integration', 'unstructured information vector conversion', 'multi-layer perceptron design', and evaluate the performance of each architecture, and confirm the proposed model based on the results. In addition, the target variables for predicting customer behavior are defined as six binary classification problems: re-purchaser, churn, frequent shopper, frequent refund shopper, high amount shopper, high discount shopper. In order to verify the usefulness of the proposed model, we conducted experiments using actual data of domestic specific online shopping company. This experiment uses actual transactions, customers, and VOC data of specific online shopping company in Korea. Data extraction criteria are defined for 47,947 customers who registered at least one VOC in January 2011 (1 month). The customer profiles of these customers, as well as a total of 19 months of trading data from September 2010 to March 2012, and VOCs posted for a month are used. The experiment of this study is divided into two stages. In the first step, we evaluate three architectures that affect the performance of the proposed model and select optimal parameters. We evaluate the performance with the proposed model. Experimental results show that the proposed model, which combines both structured and unstructured information, is superior compared to NBC(Naïve Bayes classification), SVM(Support vector machine), and ANN(Artificial neural network). Therefore, it is significant that the use of unstructured information contributes to predict customer behavior, and that CNN can be applied to solve business problems as well as image recognition and natural language processing problems. It can be confirmed through experiments that CNN is more effective in understanding and interpreting the meaning of context in text VOC data. And it is significant that the empirical research based on the actual data of the e-commerce company can extract very meaningful information from the VOC data written in the text format directly by the customer in the prediction of the customer behavior. Finally, through various experiments, it is possible to say that the proposed model provides useful information for the future research related to the parameter selection and its performance.

• Journal of Intelligence and Information Systems
• /
• v.16 no.4
• /
• pp.1-19
• /
• 2010

In offline environment, unlike traditional advertising model through TV, newspaper, and radio, online advertising model draws instantaneous responses from potential consumers and it is convenient to assess. This kind of characteristics of Internet advertising model has driven the growth of advertising model among various Internet business models. There are, conventionally classified, CPM (Cost Per Mile), CPC (Cost Per Click), and CPS (Cost Per Sales) models as Internet advertising expenditure model. These can be examined in manners regarding risks that stakeholders should stand and degree of responsibility. CPM model that is based on number of advertisement exposure is mechanically exposed to users but not actually recognized by users resulting in risk of wasted expenditure by advertisers without any advertising effect. While on aspect of media, CPS model that is based on conversion action is the most risky model because of the conversion action such as product purchase is determined by capability of advertisers not that of media. In this regard, while there are issue of CPM and CPS models disadvantageously affecting only one side of Internet advertising business model value network, CPC model has been evaluated as reasonable both to advertisers and media, and occupied the largest segment of Internet advertising market. However, CPC model also can cause fraudulent behavior such as click fraud because of the competition or dishonest amount of advertising expenditure. On the user aspect, unintentionally accessed advertisements can lead to more inappropriate expenditure from advertisers. In this paper, we suggest "CPCD"(Cost Per Coupon Download) model. This goes beyond simple clicking of advertisements and advertising expenditure is exerted when users download a coupon from advertisers, which is a concept in between CPC and CPS models. To achieve the purpose, we describe the scenario of advertiser perspective, processes, participants and their benefits of CPCD model. Especially, we suggest the new value in online coupon; "possibility of storage" and "complement for delivery to the target group". We also analyze the working condition for advertiser by a comparison of CPC and CPCD models through advertising expenditure simulation. The result of simulation implies that the CPCD model suits more properly to advertisers with medium-low price products rather than that of high priced goods. This denotes that since most of advertisers in CPC model are dealing with medium-low priced products, the result is very interesting. At last, we contemplate applicability of CPCD model in ubiquitous environment.

• The Journal of the Korea Contents Association
• /
• v.8 no.12
• /
• pp.37-44
• /
• 2008

Data entered into RFID tags are used for saving costs and enhancing competitiveness in the development of applications in various industrial areas. RFID readers perform the identification and search of hundreds of objects, which are tags. RFID technology that identifies objects on request of dynamic linking and tracking is composed of application components supporting information infrastructure. Despite their many advantages, existing applications, which do not consider elements related to real.time data communication among remote RFID devices, cannot support connections among heterogeneous devices effectively. As different network devices are installed in applications separately and go through different query analysis processes, there happen the delays of monitoring or errors in data conversion. The present study implements a RFID database handling system in semantic Web environment for integrated management of information extracted from RFID tags regardless of application. Users’ RFID tags are identified by a RFID reader mounted on an application, and the data are sent to the RFID database processing system, and then the process converts the information into a semantic Web language. Data transmitted on the standardized semantic Web base are translated by a smart browser and displayed on the screen. The use of a semantic Web language enables reasoning on meaningful relations and this, in turn, makes it easy to expand the functions by adding modules.

• Asian-Australasian Journal of Animal Sciences
• /
• v.21 no.7
• /
• pp.1003-1010
• /
• 2008

The objective of the present study was to identify differences in the expression levels of liver proteins between healthy and ketotic cows, establish a liver metabolic interrelationship of ketosis and elucidate the metabolic characteristics of the liver during ketosis. Liver samples from 8 healthy multiparous Hostein cows and 8 ketotic cows were pooled by health status and the proteins were separated by two-dimensional-electrophoresis (2D-E). Statistical analysis of gels was performed using PDQuest software 8.0. The differences in the expression levels of liver proteins (p<0.05) between ketotic and healthy cows were identified by matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF-TOF) tandem mass spectrometry. Five enzymes/proteins were identified as being differentially expressed in the livers of ketotic cows: expression of 3-hydroxyacyl-CoA dehydrogenase type-2 (HCDH), acetyl-coenzyme A acetyltransferase 2 (ACAT) and elongation factor Tu (EF-Tu) were down-regulated, whereas that of alpha-enolase and creatine kinase were up-regulated. On the basis of this evidence, it could be presumed that the decreased expression of HCDH, which is caused by high concentrations of acetyl-CoA in hepatic cells, in the livers of ketotic cows, implies reduced fatty acid ??oxidation. The resultant high concentrations of acetyl-CoA and acetoacetyl CoA would depress the level of ACAT and generate more ??hydroxybutyric acid; high concentrations of acetyl-CoA would also accelerate the Krebs Cycle and produce more ATP, which is stored as phosphocreatine, as a consequence of increased expression of creatine kinase. The low expression level of elongation factor Tu in the livers of ketotic cows indicates decreased levels of protein synthesis due to the limited availability of amino acids, because the most glucogenic amino acids sustain the glyconeogenesis pathway; thus increasing the level of alpha-enolase. Decreased protein synthesis also promotes the conversion of amino acids to oxaloacetate, which drives the Krebs Cycle under conditions of high levels of acetyl-CoA. It is concluded that the livers of ketotic cows possess high concentrations of acetyl-CoA, which through negative feedback inhibited fatty acid oxidation; show decreased fatty acid oxidation, ketogenesis and protein synthesis; and increased gluconeogenesis and energy production.