• The Journal of the Korea Contents Association
• /
• v.18 no.2
• /
• pp.231-239
• /
• 2018

In this paper, we propose a vanishing point detection method using multiple initial vanishing points. Vanishing points are important geometric information that is used for reconstructing 3D structures. Three vanishing points are detected for indoor scenes. In the previous work, it could be inaccurate to detect only one initial vanishing point, because initial vanishing point getting most highest sum of voting could be deferent from the best initial vanishing point. Therefore the method which sets multiple initial vanishing point and detects a best vanishing point from them gives us preparation for the prior case. Also in this paper, we propose a adjusting vanishing point method by postprocessing of detected vanishing points. We could detect more accurate vanishing point by using postprocessing. Experimental results show that the accuracy of the vanishing point detection is about 1~2% higher than that of the existing method through the proposed method and the performance is improved accordingly.

• The Journal of the Korea Contents Association
• /
• v.14 no.2
• /
• pp.28-38
• /
• 2014

For western music there has been a volume of researches on music information analysis for automatic transcription or content-based music retrieval. But it is hard to find the similar research on Korean traditional music. In this paper we propose several algorithms to automatically analyze the structure of Korean traditional music 'Pansori'. The proposed algorithm automatically distinguishes between the 'sound' part and 'speech' part which are named 'sori' and 'aniri', respectively, using the ratio of phonetic and pause time intervals. For rhythm called 'jangdan' classification the algorithm makes the robust decision using the majority voting process based on template matching. Also an algorithm is suggested to detect the bar positions in the 'sori' part based on Kalman filter. Every proposed algorithm in the paper works so well enough for the sample music sources of 'Pansori' that the results may be used to automatically transcribe the 'Pansori'.

• Journal of Broadcast Engineering
• /
• v.14 no.3
• /
• pp.342-356
• /
• 2009

This paper presents a vote decision-based deinterlacing scheme for false directional error correction(VDD) to convert interlaced signal into non-interlaced signal using only one fields. The VDD using the vote decision goes through four steps process. The first step extracts regions having doubt of false edge using MM-ELA method. In these regions, the edge direction is decided by the majority vote using upper adjacent pixels's information through the second step. But, we still have undecided directions, which will be decided by the majority vote and the directional average decision at the third step. This step preserves the edge directions and minimizes visual degradation. Finally, the last step interpolates undecided pixels using DOI method which can consider the fine edge direction. Although the VDD with hierarchical structure has a high complexity, it can extract delicate edge compared to other pixel-by-pixel or window-by-window deinterlacing algorithms. Simulation results show that it has significantly improved both the subjective and objective qualities of the reconstructed images.

• The Journal of the Korea Contents Association
• /
• v.9 no.8
• /
• pp.91-98
• /
• 2009

This paper presents a solution to error avalanche of deciphering where radio noise brings random bit errors in encrypted image data under ubiquitous environment. The image capturing module is to be made comprising data compression and encryption features to reduce data traffic volume and to protect privacy. Block cipher algorithms may experience error avalanche: multiple pixel defects due to single bit error in an encrypted message. The new fault tolerant scheme addresses error avalanche effect exploiting a three-dimensional data shuffling process, which disperses error bits on many frames resulting in sparsely isolated errors. Averaging or majority voting with neighboring pixels can tolerate prominent pixel defects without increase in data volume due to error correction. This scheme has 33% lower data traffic load with respect to the conventional Hamming code based approach.

• The Journal of the Korea Contents Association
• /
• v.18 no.9
• /
• pp.468-481
• /
• 2018

This study aims to empirically investigate how users' online engagement behaviors predict their uses of paid contents. To this end, the data from the 2016 Korean Media Panel Survey, which has been conducted annually by the Korea Information Society Development Institute(KISDI), were analyzed. Major findings(N=8.313) were as follows. First, the active type of online engagement(e.g., posting, commenting), which contributes to direct creation of online contents, was the most powerful predictor to explain the DV. On the other hand, relatively passive actions of user engagement(e.g., sharing, endorsing, voting) turned out to have no significant effects on the uses of paid contents, just as personality traits and online privacy concerns did. Based on these results, it is recommended that online contents or platform service providers should try to establish clearly-targeted marketing strategies, after thoroughly collecting and analyzing the data of users' various online behaviors.

• Journal of KIISE:Software and Applications
• /
• v.29 no.3
• /
• pp.192-201
• /
• 2002

Automated text classification is considered as an important method to manage and process a huge amount of documents in digital forms that are widespread and continuously increasing. Recently, text classification has been addressed with machine learning technologies such as k-nearest neighbor, decision tree, support vector machine and neural networks. However, only few investigations in text classification are studied on real problems but on well-organized text corpus, and do not show their usefulness. This paper proposes and analyzes text classification methods for a real application, email document classification task. First, we propose a combining method of multiple neural networks that improves the performance through the combinations with maximum and neural networks. Second, we present another strategy of combining multiple machine learning classifiers. Voting, Borda count and neural networks improve the overall classification performance. Experimental results show the usefulness of the proposed methods for a real application domain, yielding more than 90% precision rates.

• Journal of Nutrition and Health
• /
• v.36 no.5
• /
• pp.515-527
• /
• 2003

Objectives： A meta-analysis of the literatures was conducted to evaluate the cost-effectiveness of medical nutrition therapy by dietitians. Methods : The 30 studies were identified from a computerized search of published research on MEDLINE, Science-Direct and the PQD database until May, 2002 and a review of reference lists. The main search terms were“dietitian”,“dietary intervention”,“nutrition intervention”, “cost”,“cost-effectiveness”and“cost-benefit analysis”. The subgroup analysis was performed by publication year, study design, intervention provider, type of patient (in/out-patient) and type of cost (total cost/direct cost). Two reviewers independently selected trials for inclusion, assessed the quality and extracted the data. Results : The 30 studies were identified using the electric database search and bibliographies. The 17 trials were eligible for inclusion criteria, then the systematic review and a meta-analysis were conducted on effectiveness and cost-effectiveness of medical nutrition therapy. The quality of the studies was evaluated using the quality assessment tool for observational studies. The quality score was 0.515 $\pm$ 0.121 (range : 0.279-0.711, median : 0.466). The meta-analysis of 17 studies based on the random effect model showed that medical nutrition therapy was highly effective in treating the diseases (effect size 0.3092 : 95% confidence interval 0.2282-0.3303). The vote-counting method, one of meta-analysis methods, was applied to evaluate the cost-effectiveness of medical nutrition therapy conducted by dietitians. Two criteria (method 1, method 2) for voting were used. The calculated p-values for method 1 (more conservative method) and method 2 (less conservative method) were 0.1250 and 0.0106, respectively. Medical nutrition therapy by dietitians was significantly cost-effective in the method 2. Conclusion. This meta-analysis showed that the effectiveness of medical nutrition therapy was statistically significant in treating disease (effect size 0.3092), and that the cost-effectiveness of medical nutrition therapy was statistically significant in the method 2 (less conservative method) of vote counting. (Korean J Nutrition 36(5): 515~527, 2003)

• Communications for Statistical Applications and Methods
• /
• v.17 no.1
• /
• pp.55-66
• /
• 2010

Propensity score adjustment(PSA) has been suggested as approach to adjustment for volunteer internet survey. PSA attempts to decrease the biases arising from noncoverage and nonprobability sampling in volunteer panel internet surveys. Although PSA is an appealing method, its application for internet survey regarding Korea presidential election and its effectiveness is not well investigated. In this study, we compare the Ni Korea internet survey with the telephone survey conducted by MBMR and KBS for 2007 Korean presidential election. The result of study show that the accuracy of internet survey can be improved by using PSA. And it is critical to include covariates that highly related to the voting tendency and the role of nondemographic variables seems important to improving PSA for Korea presidential election prediction.

• Management & Information Systems Review
• /
• v.28 no.3
• /
• pp.25-44
• /
• 2009

The purpose of this paper is to examine the determinants which cause a price differential between common and preferred stock. Prior studies have shown that variables like liquidity, size, the number of outstanding shares issued can explain the price differential between common and preferred stock price. Based on year 2006 through year 2008 data, we analyzed the determinants using regression model. Dummy variables representing large/small company and KSE/KOSDAQ respectively are added and analyzed as independent variables. The firm size, trade volume turnover, and the number of preferred shares to total outstanding shares were proved to make influence on the price differential under the 5% significance level. Especially, we have found the number of preferred shares to total outstanding shares provide the most strong relationship with the price differential. This means that a high ratio of preferred stock to total outstanding shares leads to relatively high value of common stock and causes a big price differential.

• Korean Journal of Social Welfare
• /
• v.62 no.3
• /
• pp.141-162
• /
• 2010

This paper tries to provide empirical support for a formal model of social policy development which has been presented in a former paper of this study. In the direct democracy, median voter's social policy preference is critical because he is Condorcet winner in a pair-wise pure majority voting. But, in the more general setting, we should think of various political institutions as a collective choice device. For this reason, I draw a formal model which formulates three contrasting types of political institutions which are distinguished by the developments of political democracy and the differences of electoral rules. Comparative patterns of key variables which measure political institutions, social policy developments, and social policy preference provide support for my arguments. My empirical results suggest that three political institutions are associated with very different policy outcomes. Compared to other institutions, committee system entails more targeted subsidy, less universal benefit. On the contrary, proportional elections produce more universal benefit, less targeted subsidy.