• Journal of the Korea Society of Computer and Information
• /
• v.18 no.4
• /
• pp.65-77
• /
• 2013

Recently the ratio of the Internet traffic generated by video streaming applications including video-on-demand (VOD) is getting higher and higher, while P2P-based naive content distribution has been the main source of the Internet traffic in the past. As a result, the paradigm of cooperatively distributed systems (e.g., P2P) is changing to support streaming applications. Most P2P assisted approaches for video streaming today are based on Bit Torrent thanks to its simplicity of implementation and easy adaptability. They, however, have immanent vulnerability to free-riding inherited from Bit Torrent, which inevitably hurts their performance under limited system resources with free-riding. This paper studies the weakness to free-riding of existing Bit Torrent-based video streaming applications and investigates the adaptability of T-Chain (which was originally designed to prevent free-riding in cooperatively distributed systems) to video streaming applications. Our experiment results show that the video streaming approach based on T-Chain outperforms most existing Bit Torrent-based ones by 60% on average under limited system resources with free-riding.

• KIPS Transactions on Software and Data Engineering
• /
• v.5 no.5
• /
• pp.251-260
• /
• 2016

Face recognition is a technology to extract feature from a facial image, learn the features through various algorithms, and recognize a person by comparing the learned data with feature of a new facial image. Especially, in order to improve the rate of face recognition, face recognition requires various processing methods. In the training stage of face recognition, feature should be extracted from a facial image. As for the existing method of extracting facial feature, linear discriminant analysis (LDA) is being mainly used. The LDA method is to express a facial image with dots on the high-dimensional space, and extract facial feature to distinguish a person by analyzing the class information and the distribution of dots. As the position of a dot is determined by pixel values of a facial image on the high-dimensional space, if unnecessary areas or frequently changing areas are included on a facial image, incorrect facial feature could be extracted by LDA. Especially, if a camera image is used for face recognition, the size of a face could vary with the distance between the face and the camera, deteriorating the rate of face recognition. Thus, in order to solve this problem, this paper detected a facial area by using a camera, removed unnecessary areas using the facial feature area calculated via a Gabor filter, and normalized the size of the facial area. Facial feature were extracted through LDA using the normalized facial image and were learned through the artificial neural network for face recognition. As a result, it was possible to improve the rate of face recognition by approx. 13% compared to the existing face recognition method including unnecessary areas.

• Management & Information Systems Review
• /
• v.29 no.4
• /
• pp.123-142
• /
• 2010

This study suggest that firms can use online brand communities as an IMC tool to achieve high brand loyalty through marketer-controlled or loyal customer-controlled brand contacts. In this perspective, the online brand community as a marketing communication tool can help the firm in eliciting favorable responses from customers. This study finds that an online brand community, as a critical marketing promotion tool, helps a firm elicit favorable relationship with customers and build strong brand loyalty. In particular, this study suggests several important theoretical and managerial implications. First, this study confirm that "advertising usefulness" is the most powerful and important factor that affects cgerial 's positive emotionomehile "sales promotion usefulness" impacmehin "interactivity" but dies not impacmhin "cgerial iexperience"ltyevent usefulness" impacmehin "cgerial iexperience"but dies not impacmhin "interactivity." In addition, "cgerial iexperience" signifn "itly impacmehin "cgerial -to-cgerial iinteractivity." This indicates that online environment provides participapacmwith a fun and exciting environment. In that sense, enhancing the online brand community experiencemwould be a critical factor for building strong brand. Thi", mword of mouth can play a riclly important role in making many cgerial s to trust brand and to enhance online brand community loyalty. Web users are becoming web authoore owning and creating content limited only by their imaginations.

• Management & Information Systems Review
• /
• v.33 no.2
• /
• pp.225-239
• /
• 2014

An effect of reward program related to promotional activity has the limitation of being concentrated on a short-term performance or of inducing temporary re-purchase. Accordingly, this study verified the influence of reward method upon consumers' response recovery centering on interactive effects of reward timing and reward intensity, in order to expand a research of dissatisfaction situation. As for the objective of this study, first, the aim is to verify the difference in the influence of economic, non-economic, and combined rewards, which are reward methods of dissatisfaction situation, upon consumers' cognitive response recovery and emotional response recovery. Second, the aim is to confirm a moderating role of reward timing and reward intensity in the effect of consumers' response recovery according to reward methods. To design a research, the perfect factor design between subjects in 3X2X2 was used. As a result of major research, first, there was a difference in the influence upon consumers' response recovery depending on reward methods. Second, the influence of reward method upon consumers' response recovery had a difference depending on reward timing. Third, the influence of reward method upon consumers' response recovery had a difference depending on reward intensity. Consumers' response recovery level was confirmed to have the greatest influence in the combined reward. This study has a significance in newly applying the reward timing, in the dissatisfaction situation which is addressed in the general reward program. Through this study, the aim was to support the empirically analytical results of prior researches and to expand its role in several angles.

• Weed & Turfgrass Science
• /
• v.3 no.4
• /
• pp.262-268
• /
• 2014

This study was conducted to provide basic information for weed control by surveying the occurrence of weed species. Surveys of weed species occurred in food crop (barley, maize, potato, soybean) fields were conducted in Chungnam province from April to September in 2014. Total 321 sites of food crop fields in 17 City / Gun in Chungnam province were investigated. From the result of this survey, 130 weed species in 36 families were identified and classified to 77 annuals, 27 biennials and 26 perennials. Based on the occurrence ratio, the most dominant weed species in Chungnam province food crop fields were Digitaria ciliaris (10.19%), followed by Conyza canadensis (7.82%), Portulaca oleracea (5.22%), Chenopodium album var. centrorubrum (5.21%), Capsella bursa-pastoris (3.47%) and Alopecurus aequalis (3.47%). Forty-four exotic weed species were identified. This information could be useful for estimation of future weed occurrence, weed population dynamics and establishment of weed control methods in food crop fields of Chungnam province.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.13 no.2
• /
• pp.615-627
• /
• 2012

The market penetration of the smart phones has brought significant changes to the related industries. As the mobile phone market as well as the smart phone application market are growing rapidly, competition among small-size application developers has become severe. However, due to the severe competition and expensive market entry costs, the developers argue that it is necessary to develope the applications from the perspective of the users. However, studies on application development from the users' perspectives are still in the early stages and they have not provided various approaches. Therefore, based on the Open Coding method of Ground Theory, this study collected data on applications review from related communities and blogs of Korean web portal sites, and identified indices which users consider important when they use the applications. In addition, we conduct a comparative analysis between those indices by calculating their frequency of exposure. As a result, a total of 30 sub-categories of indicators such as amusement, controllability, versatility and ease of use appeared to be predominant to users and those lower categories were grouped into five categories; sensibility, design, technology, price, and social skills. The results of this study suggest to the application developers the guidelines of user oriented design and development. It can be used to develop the evaluation tool for application usability.

• Journal of Korean Society for Geospatial Information Science
• /
• v.21 no.4
• /
• pp.73-81
• /
• 2013

Ordinary least squares (OLS) regression is the primary statistical method in previous studies for vegetation distribution patterns in relation to landform. However, this global regression lacks the ability to uncover some local-specific relationships and spatial autocorrelation in model residuals. This study employed geographically weighted regression (GWR) to examine the spatially varying relationships between NDVI (Normalized Difference Vegetation Index) patterns and changing trends of landform (elevation, slope) and solar intensity (insolation and duration of sunshine) in Mt Geum-gang of North-Korea. Results denoted that GWR was more powerful than OLS in interpreting relationships between NDVI patterns and landform/solar intensity, since GWR was characterized by higher adjusted R2, and reduced spatial autocorrelations in model residuals. Unlike OLS regression, GWR allowed the coefficients of explanatory variables to differ by locality by giving relatively more weight to NDVI patterns which are affected by local landform and solar factors. The strength of the regression relationships in the GWR increased significantly, by showing regression coefficient of higher than 70% (0.744) in the southern ridge of the experimental area. It is anticipated that this research output will serve to increase the scientific and objective vegetation monitoring in relation to landform and solar intensity by overcoming serious constraints suffered from the past non-GWR-based approach.

• Spatial Information Research
• /
• v.22 no.4
• /
• pp.77-87
• /
• 2014

Recently there are many disasters caused by volcanic activities such as the eruptions in Tungurahua, Ecuador(2014) and $Eyjafjallaj\ddot{o}kull$, Iceland(2010). Therefore, it is required to prepare countermeasures for the disasters. This study analyzes the Baekdu Mountain area, where is the risky area because it is active volcano, based on the observed data and scientific methods in order to assess a risk, produce a hazard map and analyze a degree of risk caused by the volcano. Firstly, it is reviewed for the research about the Baekdu mountain volcanic eruption in 1215(${\pm}15$ years) done by Liu Ruoxin. And the factors causing volcanic disaster, environmental effects, and vulnerability of Baekdu Mountain are assessed by the dataset, which includes the earthquake monitoring data, the volcanic deformation monitoring data, the volcanic fluid geochemical monitoring data, and the socio-economic statistics data. A hazard, especially caused by a volcano, distribution map for the Baekdu Mountain Area is produced by using the assessment results, and the map is used to establish the disaster risk index system which has the four phases. The first and second phases are very high risky area when the Baekdu Mountain erupts, and the third and fourth phases are less dangerous area. The map shows that the center of mountain has the first phase and the farther area from the center has the lower phase. Also, the western of Baekdu Mountain is more vulnerable to get the risk than the eastern when the factors causing volcanic disasters are equally applied. It seems to be caused by the lower stability of the environment and the higher vulnerability.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.21 no.6
• /
• pp.679-686
• /
• 2020

In this thesis, an enhanced method for the feature extraction of vocal source signals and score combination using an MCE-Based weight estimation of the score of multiple feature vectors are proposed for the performance improvement of speaker recognition systems. The proposed feature vector is composed of perceptual linear predictive cepstral coefficients, skewness, and kurtosis extracted with lowpass filtered glottal flow signals to eliminate the flat spectrum region, which is a meaningless information section. The proposed feature was used to improve the conventional speaker recognition system utilizing the mel-frequency cepstral coefficients and the perceptual linear predictive cepstral coefficients extracted with the speech signals and Gaussian mixture models. In addition, to increase the reliability of the estimated scores, instead of estimating the weight using the probability distribution of the convectional score, the scores evaluated by the conventional vocal tract, and the proposed feature are fused by the MCE-Based score combination method to find the optimal speaker. The experimental results showed that the proposed feature vectors contained valid information to recognize the speaker. In addition, when speaker recognition is performed by combining the MCE-based multiple feature parameter scores, the recognition system outperformed the conventional one, particularly in low Gaussian mixture cases.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.14 no.1
• /
• pp.1-11
• /
• 2011

The objective of this study was to investigate the landscape structures of the National and Provincial Forest Parks in the list of "100 Famous Forests of Korea" designated by Korea Forest Service by comparing them with those of the non-designated forests. Among the designated and the non-designated, 34 mountain forests were chosen respectively over all provinces of Korea. The spatial characteristics of forest landscapes were quantified as the landscape indices independently using FRAGSTATS and the two sets of results of analyses were compared each other. According to the results of the comparative study, the designated forests were found on the higher elevation with the higher average slope and the more complicated relief conditions rather than those of the non-designated. In terms of landscape structure, the designated forests show the larger average patch size, the lower edge density and the higher diversity of landscape components. These results indicate that the more hilly mountain forests with the more complicated spatial distribution patterns of patches are the characteristics of the designated forests. The indices of the forest landscape structure would be useful in understanding the perception of forest landscape.