• Journal of Korea Society of Industrial Information Systems
• /
• v.12 no.4
• /
• pp.15-24
• /
• 2007

The recognition of postal address images is indispensable for the automatic sorting of postal envelopes. The process of the address image recognition is composed of three steps-address image preprocessing, character recognition, address interpretation. The extracted character images from the preprocessing step are forwarded to the character recognition step, in which multiple candidate characters with reliability scores are obtained for each character image extracted. aracters with reliability scores are obtained for each character image extracted. Utilizing those character candidates with scores, we obtain the final valid address for the input envelope image through the address interpretation step. The envelope sorting rate depends on the performance of all three steps, among which character recognition step could be said to be very important. The good character recognizer would be the one which could produce valid candidates with very reliable scores to help the address interpretation step go easy. In this paper, we propose the method of generating character candidates with reliable recognition scores. We utilize the existing MLP(multilayered perceptrons) neural network of the address recognition system in the current automatic postal envelope sorters, as the classifier for the each image from the preprocessing step. The MLP is well known to be one of the best classifiers in terms of processing speed and recognition rate. The false alarm problem, however, might be occurred in recognition results, which made the address interpretation hard. To make address interpretation easy and improve the envelope sorting rate, we propose promising methods to reestimate the recognition score (confidence) of the existing MLP classifier: the generation method of the statistical recognition properties of the classifier and the method of the combination of the MLP and the subspace classifier which roles as a reestimator of the confidence. To confirm the superiority of the proposed method, we have used the character images of the real postal envelopes from the sorters in the post office. The experimental results show that the proposed method produces high reliability in terms of error and rejection for individual characters and non-characters.

• Journal of Intelligence and Information Systems
• /
• v.19 no.2
• /
• pp.39-54
• /
• 2013

Recent explosive increase of electronic commerce provides many advantageous purchase opportunities to customers. In this situation, customers who do not have enough knowledge about their purchases, may accept product recommendations. Product recommender systems automatically reflect user's preference and provide recommendation list to the users. Thus, product recommender system in online shopping store has been known as one of the most popular tools for one-to-one marketing. However, recommender systems which do not properly reflect user's preference cause user's disappointment and waste of time. In this study, we propose a novel recommender system which uses data mining and multi-model ensemble techniques to enhance the recommendation performance through reflecting the precise user's preference. The research data is collected from the real-world online shopping store, which deals products from famous art galleries and museums in Korea. The data initially contain 5759 transaction data, but finally remain 3167 transaction data after deletion of null data. In this study, we transform the categorical variables into dummy variables and exclude outlier data. The proposed model consists of two steps. The first step predicts customers who have high likelihood to purchase products in the online shopping store. In this step, we first use logistic regression, decision trees, and artificial neural networks to predict customers who have high likelihood to purchase products in each product group. We perform above data mining techniques using SAS E-Miner software. In this study, we partition datasets into two sets as modeling and validation sets for the logistic regression and decision trees. We also partition datasets into three sets as training, test, and validation sets for the artificial neural network model. The validation dataset is equal for the all experiments. Then we composite the results of each predictor using the multi-model ensemble techniques such as bagging and bumping. Bagging is the abbreviation of "Bootstrap Aggregation" and it composite outputs from several machine learning techniques for raising the performance and stability of prediction or classification. This technique is special form of the averaging method. Bumping is the abbreviation of "Bootstrap Umbrella of Model Parameter," and it only considers the model which has the lowest error value. The results show that bumping outperforms bagging and the other predictors except for "Poster" product group. For the "Poster" product group, artificial neural network model performs better than the other models. In the second step, we use the market basket analysis to extract association rules for co-purchased products. We can extract thirty one association rules according to values of Lift, Support, and Confidence measure. We set the minimum transaction frequency to support associations as 5%, maximum number of items in an association as 4, and minimum confidence for rule generation as 10%. This study also excludes the extracted association rules below 1 of lift value. We finally get fifteen association rules by excluding duplicate rules. Among the fifteen association rules, eleven rules contain association between products in "Office Supplies" product group, one rules include the association between "Office Supplies" and "Fashion" product groups, and other three rules contain association between "Office Supplies" and "Home Decoration" product groups. Finally, the proposed product recommender systems provides list of recommendations to the proper customers. We test the usability of the proposed system by using prototype and real-world transaction and profile data. For this end, we construct the prototype system by using the ASP, Java Script and Microsoft Access. In addition, we survey about user satisfaction for the recommended product list from the proposed system and the randomly selected product lists. The participants for the survey are 173 persons who use MSN Messenger, Daum Caf$\acute{e}$, and P2P services. We evaluate the user satisfaction using five-scale Likert measure. This study also performs "Paired Sample T-test" for the results of the survey. The results show that the proposed model outperforms the random selection model with 1% statistical significance level. It means that the users satisfied the recommended product list significantly. The results also show that the proposed system may be useful in real-world online shopping store.

• Journal of Pharmaceutical Investigation
• /
• v.38 no.4
• /
• pp.269-275
• /
• 2008

The purpose of the present study was to evaluate the bioequivalence of two levofloxacin tablets, Jeil $Cravit^{TM}$ tablet (Jeil Pharm. Co., Ltd., Korea, reference drug) and $Lesacin^{TM}$ tablet (Ilhwa. Co., Ltd., Korea, test drug), according to the guidelines of Korea Food and Drug Administration (KFDA). Twenty-four healthy male Korean volunteers received two tablets containing levofloxacin 200 mg in a $2{\times}2$ crossover study. There was a one-week washout period between the doses. Plasma concentrations of levofloxacin were monitored for over a period of 24 hr after administration by using a high performance liquid chromatography-tandem mass spectrometry (LC-MS/MS). The area under the plasma concentration-time curve from time zero to 24 hr ($AUC_t$), maximum plasma drug concentration ($C_{max}$) and time to reach $C_{max}\;(T_{max})$ were complied from the plasma concentration-time data. Analysis of variance (ANOVA) test was utilized for the statistical analysis of the parameters using logarithmically transformed $AUC_t$ and $C_{max}$. The 90% confidence intervals of the $AUC_t$ ratio and the $C_{max}$ ratio for $Lesacin^{TM}$/Jeil $Cravit^{TM}$ were $\log\;0.9527{\sim}\log\;0.9981$ and $\log\;0.8712{\sim}\log\;1.0556$, respectively. These values were within the acceptable bioequivalence intervals of $\log\;0.80{\sim}\log\;1.25$, recommended by KFDA. In all of these results, we concluded that $Lesacin^{TM}$ tablet was bioequivalent to Jeil $Cravit^{TM}$ tablet, in terms of rate and extent of absorption.

• Journal of Pharmaceutical Investigation
• /
• v.38 no.4
• /
• pp.261-267
• /
• 2008

The aim of the present study was to evaluate the bioequivalence of two gabapentin preparations. We used Neurontin tablet 800 mg (Pfizer Korea Inc.) as a reference drug for bioequivalence of Gabalep tablet 800 mg (Chong Kun Dang Pharmaceutical Co., Korea), and performed this whole study according to the guidelines of Korea Food and Drug Administration (KFDA). Twenty five healthy male volunteers were administered with each drug in a randomized $2{\times}2$ cross-over study with one week washout interval. After drug administration, blood was taken at predetermined time intervals ($0{\sim}24$ hours) and the concentrations of gabapentin in serum were determined using an high performance liquid chromatography-tandem mass spectrometer (LC-MS/MS) employing electrospray ionization technique and operating in multiple reaction mornitoring (MRM). The analytical method was validated in specificity, accuracy, precision and linearity. The phar-macokinetic parameters such as AUCt and Cmax were calculated and ANOVA test was utilized for the statistical analysis of the parameters using logarithmically transformed AUCt and Cmax. $Mean{\pm}SD$. of AUCt and Cmax value for reference drug and test drug were $29.94{\pm}9.23\;({\mu}g/mL{\cdot}hr)$ and $3.12{\pm}1.11\;({\mu}g/mL{\cdot}hr)$, and $31.48{\pm}9.77\;({\mu}g/mL{\cdot}hr)$ and $3.15{\pm}1.03\;({\mu}g/mL)$, respectively. The 90% confidence intervals using logarithmically transformed data were within the acceptance range of log(0.8) to log(1.25) for AUCt and Cmax, respectively. These results indicate that Gabalep tablet 800 mg is bioequivalent to Neurontin tablet 800 mg.

• The Journal of the Korea Contents Association
• /
• v.9 no.4
• /
• pp.265-274
• /
• 2009

This study aims at examining the actual condition of recognition of dental caries(ACRDC), presenting a scheme to improve the recognition, and providing basic data necessary to prevent oral diseases, Dental caries (DC) which one of the three most important chronic diseases in Korea. A survey was conducted on ACRDC for about 390 people twenty and over who visited dental clinics for dental treatment 336 questionnaires with exception of omitted answer were used for statistical analysis. Many of them had never heard of DC and most of them failed to recognize it. People who experienced DC had a dental clinic as a main information channel and they were significant differences in those obtaining information through other channels(p=0.008). Univariate logistic regression analysis for effects on ACRDC demonstrated that possibility of subjects in their twenties recognizing DC more accurately than those in their forties Odds ratio(95% confidence Interval) was 2.06(1.16-3.66)(p=0.000), with the possibility of professionals recognizing the disease more accurately than those with other kinds of occupation 5.49(2.52-11.93)(p=0.000), showing statistically significant relevance. As for medical security, possibility of work insurance being related to more accurate recognition of dental caries than others was 1.95 (1.03-3.71)(p=0.003), with the possibility of subjects whose dental state was very good recognizing dental caries more accurately than those whose dental state was very bad 6.40(1.57-26.03) (p=0.002), showing statistically significant relevance. For prevention of DC, an education through experts working at a dental clinic are necessary for those in their twenties and over, who were other than professionals, who were insured for medical security via other than work insurance, and whose dental state was bad.

• Korean Journal of Environment and Ecology
• /
• v.6 no.2
• /
• pp.162-167
• /
• 1993

A mixed deciduous forest community in Mt. Sobaek was studied to determine the adequate number of plots of tree stratum for investigating forest community structure. Twenty l0m $\times$ l0m plots were set up iii the studied forest community, and species area curve. performance curve and statistical method were carried out. According to species-area curve, the minimal number of plots where a given percentage increase in number of plots produced less than the same percentage increase in number of species was eight. The minimal number of plots where a given percentage increase in number of plots produced less than the half of the percentage increase in number of plots was eleven. According to performance curve by importance value of the major species, the minimal number of plots where the dominant species was distinguished from the subdominant species was five. The minimal number of plots where the subdominant species was distinguished from each other was ten. Therefore, ten l0m $\times$ l0m plots seems to give an adequate sample for investigating structure of the studied forest community. Similarity index between the ten plots and total twenty plots was above 90%, and 95% confidence interval of species diversity of the ten plots was $\pm$ 0.073.

• Journal of Distribution Science
• /
• v.12 no.4
• /
• pp.23-30
• /
• 2014

Purpose - The purpose of this study is to analyze the efficiency of retail businesses by dividing domestic retailers into discount stores, super supermarkets (SSMs), and department stores. It suggests retail-business investment strategies by using data environment analysis (DEA) to analyze how input elements such as store area, parking lot area, number of employees, and sales management expenses for the convenience of customers positively affect business performance measurements such as sales and visiting customers per day. Research Design, Data, and Methodology - The DEA model calculates a ratio of the weighted mean of various inputs to the weighted mean of various outputs and measures the efficiency of a specific decision making unit (DMU). The study included 19 companies (five discount store DMUs, ten SSM DMUs, and four department store DMUs). Because the business elements and sizes of retail store DMUs used in this analysis are different, average per-store input and output variables were used. Data were collected from "The Yearbook of Retail Industry in Korea (2012)." DEA analysis was used to determine differences in efficiency among discount stores, SSMs, and department stores in terms of the business elements of each retail business. It was also used to determine what business elements were excessively invested in by comparing and analyzing efficiency by business elements using SPSS software's ANOVA (Analysis of Variance). Results - The CCR and BCC efficiency analysis found that the efficiency of discount stores is low. We believe that the saturation state of discount stores is a major factor. The ANOVA analysis confirms the VRS hypothesis with a statistically significant difference among the three groups, based on an analysis confidence interval of 95%. CRS and SE were not found to be significantly different among the three groups. As for the post hoc test, which concretely shows differences by group, the Scheffe's multiple comparison analysis test found the average differences between group 1 (discount stores) and group 2 (SSM) to be statistically significant. Conclusions - The DEA efficiency analysis implies that investment in input elements, including store area, parking lot area, and sales management expenses, were excessive in the case of discount stores, while SSMs need to invest more in promotion activities such as gifts, events, and coupons for customer management. Department stores have found that small companies invest excessively in input elements. Department stores need to invest in differentiated shopping mall complexes. This study was limited in acquiring statistical data; various input variables which might have shown more secure customer management and promotional expenses could not be applied. As the study was limited in various aspects of the efficiency analyses because financial analyses of the companies and of causal relationships, including satisfaction and loyalty of visiting customers, were not done, these aspects will be examined in the next study.

• Asian Pacific Journal of Cancer Prevention
• /
• v.13 no.11
• /
• pp.5773-5775
• /
• 2012

Objective: To assess the diagnostic and prognostic value of AFP and des-gamma-carboxyprothrombin (DCP) in combination and alone for hepatocellular carcinoma. Materials and Methods: A case control study carried out in the Department of Biochemistry of Manipal College of Medical Sciences, Pokhara, Nepal between $1^{st}$ January 2010 and $31^{st}$ December 2011. The variables collected were age, gender, BMI, total proteins, albumin, AST, ALT, total bilirubin, DCP, AFP. Approval for the study was obtained from the institutional research ethical committee. Estimation of AFP was performed by ELISA reader for all cases. Analysis was done using descriptive statistics and confidence interval (CI). The data was analyzed using Excel 2003, R 2.8.0 Statistical Package for the Social Sciences (SPSS) for Windows Version 16.0 (SPSS Inc; Chicago, IL, USA) and the EPI Info 3.5.1 Windows Version. Results:The mean age of HCC cases was $53.6{\pm}14.93$ yrs. The percentage of females was less than males in both cases (23%) and controls (29%). The specificity of DCP reached 100% when its values was equal or greater than 150 (MAU/ml) for 0, 3, 6, 9, 12 months preceding the diagnosis of HCC. Similarly, the specificity for AFP was also nearly 100% when its value was equal or greater than 200 ng/ml 0, 3, 6, 9, 12 months earlier to the finding of HCC. The specificity of DCP (${\geq}40MAU/mL$) and AFP(${\geq}20$ ng/mL) in combination was 93%, 97%, 95%, 96%, 97% in respect to 0, 3, 6, 9, 12 months prior to the diagnosis of HCC. Conclusion: The combination of both DCP and AFP will improve the finding of initial HCC and the sensitivity of these markers was utmost at the time of HCC identification and noticeably lesser at former time points.

• Asian Pacific Journal of Cancer Prevention
• /
• v.15 no.19
• /
• pp.8301-8310
• /
• 2014

Background: Published data regarding associations between the -765G>C polymorphism in cyclooxygenase-2 (COX-2) gene and digestive system cancer risk have been inconclusive. The aim of this study was to comprehensively evaluate the genetic risk of the -765G>C polymorphism in the COX-2 gene for digestive system cancer. Materials and Methods: A search was performed in Pubmed, Medline (Ovid), Embase, CNKI, Weipu, Wanfang and CBM databases, covering all studies until Feb 10, 2014. Statistical analysis was performed using Revman5.2. Results: A total of 10,814 cases and 16,174 controls in 38 case-control studies were included in this meta-analysis. The results indicated that C allele carriers (GC+CC) had a 20% increased risk of digestive system cancer when compared with the homozygote GG (odds ratio (OR)=1.20, 95% confidence interval (CI), 1.00-1.44 for GC+CC vs GG). In the subgroup analysis by ethnicity, significant elevated risks were associated with C allele carriers (GC+CC) in Asians (OR = 1.46, 95% CI=1.07-2.01, and p=0.02) and Africans (OR=2.12, 95% CI=1.57-2.87, and p< 0.00001), but not among Caucasians, Americans and mixed groups. For subgroup analysis by cancer type (GC+CC vs GG), significant associations were found between the -765G>C polymorphism and higher risk for gastric cancer (OR=1.64, 95% CI=1.03-2.61, and p=0.04), but not for colorectal cancer, oral cancer, esophageal cancer, and others. Regarding study design (GC+CC vs GG), no significant associations were found in then population-based case-control (PCC), hospital-based case-control (HCC) and family-based case-control (FCC) studies. Conclusions: This meta-analysis suggested that the -765G>C polymorphism of the COX-2 gene is a potential risk factor for digestive system cancer in Asians and Africans and gastric cancer overall.

• Journal of Korean Academy of Nursing
• /
• v.29 no.5
• /
• pp.1087-1102
• /
• 1999

Diabetes mellitus, a universal and prevalent chronic disease, is projected to be one of the most formidable worldwide health problems in the 21st century. For those living with diabetes, there is a need for self-care skills to manage a complex medical regimen. Self-efficacy which refers to one's belief in his/her capability to monitor and perform the daily activities required to manage diabetes has be found to be related to self-care. The concept of self-efficacy comes from social cognitive theory which maintains that cognitive mechanism mediate the performance of behavior. The literature cites several research studies which show a strong relationship between self-efficacy and self-care behavior. Meta-analysis is a technique that enables systematic review and quantitative integration of the results from multiple primary studies that are relevant to a particular research question. Therefore, this study was done using meta-analysis to quantitatively integrate the results of independent research studies to obtain numerical estimates of the overall effect of a self-efficacy with diabetic patient on self-care behaviors. The research proceeded in three stages : 1) literature search and retrieval of studies in which self-efficacy was related to self-care, 2) coding, and 3) calculation of mean effect size and data analysis. Seventeen studies which met the research criteria included study population of adults with diabetes, measures of self-care and measures of self-efficacy as a predictive variable. Computation of effect size was done on DSTAT which is a statistical computer program specifically designed for meta-analysis. To determine the effect of self-efficacy on self-care practice homogeneity tests were conducted. Pooled effect size estimates, to determine the best subvariable for composite variables, metabolic control variables and component of self-efficacy and self-care, indicated that the effect of self-efficacy composite on self-care composite was moderate to large. The weighted mean effect size of self-efficacy composite and self-care composite were ＋.76 and the confidence interval was from ＋.66 to ＋.86 with the number of subjects being 1,545. The total for this meta-analysis result showed that the weighted mean effect sizes ranged from ＋.70 to ＋1.81 which indicates a large effect. But since reliabilities of the instruments in the primary studies were low or not stated, caution must be applied in unconditionally accepting the results from these effect sizes. Meta-analysis is a useful took for clarifying the status of knowledge development and guiding decision making about future research and this study confirmed that there is a relationship between self-efficacy and self-care in patients with diabetes. It, thus, provides support for nurses to promote self-efficacy in their patients. While most of the studies included in this meta-analysis used social cognitive theory as a framework for the study, some studies use Fishbein & Ajzen's attitude model as a model for active self-care. Future research is needed to more fully define the concept of self-care and to determine what it is that makes patients feel competent in their self-care activities. The results of this study showed that self-efficacy can promote self-care. Future research is needed with experimental design to determine nursing interventions that will increase self-efficacy.