• 대한가정학회지
• /
• 제45권2호
• /
• pp.91-103
• /
• 2007

The purpose of this study is to investigate the types and functions of brand labels on clothing. We surveyed the materials and manufacturing methods for brand labels by visiting the label stores and label manufacturers. 200 pieces of children's wear were surveyed. The label attributes that were studied were: the number of labels, the location of the labels, the attachment system for the labels, the color of the labels, the materials used to make the labels, manufacturing methods, and the size of the labels. From this investigation a brand label was classified into a main label and a point label. The main results were: 1. Materials such as fabrics, nonwovens, leather, suede, rubber, PVC, silicone, and metals are used for brand labels. The manufacturing methods for brand labels are weaving, printing, high frequency, heating, and molding. 2. More than 54% of clothes have more than two brand labels attached. This percentage exceeds the attaching of only one brand label in rate. An inside brand label is located at a certain place. This inside label uses only fabric material reflecting inherent brand color and design. The outside brand label is located at several places with consideration of the clothes design. This label uses various materials, colors, and characters matching with the clothes. As for the size, an inside label is mainly medium in size, whereas an outside label is small. 3. A brand label is classified into a main label (first label) and a point label (second label), which are defined as follows. A main label indicates the brand name and is located inside at a certain place using an inherent brand design and a fabric material. A point label is an additional label to express brand image and is located outside at various places for decoration using various characters and design and materials.

• 한국컴퓨터정보학회논문지
• /
• 제23권6호
• /
• pp.1-7
• /
• 2018

A multi-label classification is to find multiple labels associated with the input pattern. Multi-label classification can be achieved by extending conventional single-label classification. Common extension techniques are known as Binary relevance, Label powerset, and Classifier chains. However, most of the extended multi-label naive bayes classifier has not been able to accurately estimate posterior probabilities because it does not reflect the label dependency. And the remaining extended multi-label naive bayes classifier has a problem that it is unstable to estimate posterior probability according to the label selection order. To estimate posterior probability well, we propose a new posterior probability estimation method that reflects the probability between all labels and labels efficiently. The proposed method reflects the correlation between labels. And we have confirmed through experiments that the extended multi-label naive bayes classifier using the proposed method has higher accuracy then the existing multi-label naive bayes classifiers.

• 지능정보연구
• /
• 제27권3호
• /
• pp.175-197
• /
• 2021

최근 딥 러닝 기술의 발전으로 뉴스, 블로그 등 다양한 문서에 포함된 텍스트 분석에 딥 러닝 기술을 활용하는 연구가 활발하게 수행되고 있다. 다양한 텍스트 분석 응용 가운데, 텍스트 분류는 학계와 업계에서 가장 많이 활용되는 대표적인 기술이다. 텍스트 분류의 활용 예로는 정답 레이블이 하나만 존재하는 이진 클래스 분류와 다중 클래스 분류, 그리고 정답 레이블이 여러 개 존재하는 다중 레이블 분류 등이 있다. 특히, 다중 레이블 분류는 여러 개의 정답 레이블이 존재한다는 특성 때문에 일반적인 분류와는 상이한 학습 방법이 요구된다. 또한, 다중 레이블 분류 문제는 레이블과 클래스의 개수가 증가할수록 예측의 난이도가 상승한다는 측면에서 데이터 과학 분야의 난제로 여겨지고 있다. 따라서 이를 해결하기 위해 다수의 레이블을 압축한 후 압축된 레이블을 예측하고, 예측된 압축 레이블을 원래 레이블로 복원하는 레이블 임베딩이 많이 활용되고 있다. 대표적으로 딥 러닝 모델인 오토인코더 기반 레이블 임베딩이 이러한 목적으로 사용되고 있지만, 이러한 기법은 클래스의 수가 무수히 많은 고차원 레이블 공간을 저차원 잠재 레이블 공간으로 압축할 때 많은 정보 손실을 야기한다는 한계가 있다. 이에 본 연구에서는 오토인코더의 인코더와 디코더 각각에 스킵 연결을 추가하여, 고차원 레이블 공간의 압축 과정에서 정보 손실을 최소화할 수 있는 레이블 임베딩 방법을 제안한다. 또한 학술연구정보서비스인 'RISS'에서 수집한 학술논문 4,675건에 대해 각 논문의 초록으로부터 해당 논문의 다중 키워드를 예측하는 실험을 수행한 결과, 제안 방법론이 기존의 일반 오토인코더 기반 레이블 임베딩 기법에 비해 정확도, 정밀도, 재현율, 그리고 F1 점수 등 모든 측면에서 우수한 성능을 나타냄을 확인하였다.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 1999년도 추계종합학술대회 논문집
• /
• pp.109-112
• /
• 1999

In this paper, we propose more effective method of label allocation on Multi-Protocol Label Switching (MPLS) which is IP over ATM integrated model. We research the problems, one is using downstream label allocation method case, the other is using both downstream and upstream label allocation method. Easily we can solve this problem through the downstream-on-demand label allocation method with RSVP(Resource ReSerVation Protocol). In experiment we can find 1.5~28% error which will be fixed by using downstream-on-demand label allocation method.

• 한국컴퓨터정보학회논문지
• /
• 제23권9호
• /
• pp.7-13
• /
• 2018

Recent advancement in data gathering technique improves the capability of information collecting, thus allowing the learning process between gathered data patterns and application sub-tasks. A pattern can be associated with multiple labels, demanding multi-label learning capability, resulting in significant attention to multi-label feature selection since it can improve multi-label learning accuracy. However, existing evolutionary multi-label feature selection methods suffer from ineffective search process. In this study, we propose a evolutionary search process for the task of multi-label feature selection problem. The proposed method creates large set of offspring or new feature subsets and then retains the most promising feature subset. Experimental results demonstrate that the proposed method can identify feature subsets giving good multi-label classification accuracy much faster than conventional methods.

• Nutrition Research and Practice
• /
• 제9권1호
• /
• pp.63-70
• /
• 2015

BACKGROUND/OBJECTIVES: Use of nutrition labels in food selection is recommended for consumers. The aim of this study is to examine factors, mainly beliefs explaining nutrition label use in female college students based on the Theory of Planned Behavior (TPB). SUBJECTS/METHODS: The subjects were female college students from a university in Seoul, Korea. The survey questionnaire was composed of items examining general characteristics, nutrition label use, behavioral beliefs, normative beliefs, corresponding motivation to comply, and control beliefs. The subjects (n = 300) responded to the questionnaire by self-report, and data from 275 students were analyzed using t-test or ${\chi}^2$-test. RESULTS: The results showed that 37.8% of subjects were nutrition label users. Three out of 15 behavioral beliefs differed significantly by nutrition label use. Nutrition label users agreed more strongly on the benefits of using nutrition labels including 'comparing and selecting better foods' (P < 0.001), 'selecting healthy foods' (P < 0.05). The negative belief of 'annoying' was stronger in non-users than in users (P < 0.001). Three out of 7 sources (parents, siblings, best friend) were important in nutrition label use. Twelve out of 15 control beliefs differed significantly by nutrition label use. These included beliefs regarding constraints of using nutrition labels (e.g., time, spending money for healthy foods) and lack of nutrition knowledge (P < 0.001). Perceived confidence in understanding and applying the specifics of nutrition labels in food selection was also significantly related to nutrition label use (P < 0.001). CONCLUSIONS: This study found that the beliefs, especially control beliefs, suggested in the TPB were important in explaining nutrition label use. To promote nutrition label use, nutrition education might focus on increasing perceived control over constraints of using nutrition labels, acquiring skills for checking nutrition labels, as well as the benefits of using nutrition labels and receiving support from significant others for nutrition label use.

• 유통과학연구
• /
• 제17권5호
• /
• pp.95-102
• /
• 2019

Purpose - The purpose of the research is to identify the demographic characteristics of the customers with high private label purchase intention. According to the previous research demographics such as gender, age, income, and residence type affect private label purchase intention indirectly through psychographics rather than directly. For instance, higher income group is time pressured, price-insensitive, quality-sensitive, less likely to enjoy shopping utilitarian products, and less likely to be variety-seeking. The main contribution of this research is to verify the results found in the previous empirical foreign research using scanner data and to investigate the differences of the characteristics of private label users between Korea and the foreign countries. Research design, data, and methodology - In order to empirically test the proposed hypotheses, scanner data of a Korean major super center was analyzed. Results - Empirical results show that private labels are more favored by old people over 50s, dwellers in individual house, lower income group, and frequent store visitors. Age of 30s, dwellers in the apartment of 30 pyung, higher income group, and consumers who purchased a large amount are less likely to purchase private labels. Gender turned out not to affect private label purchase. It should be noted that there is a significant multicollinearity among independent variables. Conclusions - The research findings provide managerial implication for retailers' private label strategy. In general, retailers heavily send private label coupons to the customers with high purchase volume. According to the research, however, store visit frequency is much more positively associated with private label purchase than purchase amount. The study has some limitations. The samples are only consumers with private label purchase experience. The data were drawn from one store and only 8 commodity products were used for the analysis. Also, if more demographics were available, a more complete description on the private brand users' profile could have been derived. We propose the following future research. Research using the data including consumers without private label experience, research investigating direction of causality between private label loyalty and store loyalty, and research using hedonic private label products such as TV and PC could be promising.

• 약학회지
• /
• 제58권2호
• /
• pp.112-124
• /
• 2014

This study aimed to develop a regulation system for off-label drug use to secure the safe use of marketed drugs. We searched governmental documents for national and global regulating systems of off-label drug uses and a body of academic literature to explore current regulating trends. We included European Union, United Kingdom, United States of America, Australia and Japan, and critically reviewed the regulation of off-label drug use in four issues, which were a regulatory structure, safety control before and after off-label use, and information management. The findings of the present investigation called for several measures in off-label drug uses: enhancing prescribers' self-regulation, providing up-to-date information to prescribers for evidence-based practice and to patients for their informed consent, making evidence with scientific rigor, building an official registering process for off-label use in good quality and extending the role of pharmaceutical industry in pharmacovigilance. At last, we proposed a new system so as to regulate and evaluate off-label drug uses both at national and institutional level. In the new system, we suggested a clear-cut definition for clinical evidence that applicants would submit. We newly introduced an official 'Off-Label Drug Use Report' to evaluate the safety and clinical efficacy of a given off-label drug use. In addition, we developed an algorism of the regulation of off-label drug use within an institution to help set up the culture of evidence-based practices in off-label drug uses.

• International Journal of Contents
• /
• 제6권1호
• /
• pp.6-11
• /
• 2010

Recently, there has been research to use portable digital camera to recognize objects in natural scene images, including labels or marks on a cylindrical surface. In many cases, text or logo in a label can be distorted by a structural movement of the object on which the label resides. Since the distortion in the label can degrade the performance of object recognition, the label should be rectified or restored from deformations. In this paper, a new method for label detection and restoration in digital images is presented. In the detection phase, the Hough transform is employed to detect two vertical boundaries of the label, and a horizontal edge profile is analyzed to detect upper-side and lower-side boundaries of the label. Then, the biquadratic transformation is used to restore the rectangular shape of the label. The proposed algorithm performs restoration of 3D objects in a 2D space, and it requires neither an auxiliary hardware such as 3D camera to construct 3D models nor a multi-camera to capture objects in different views. Experimental results demonstrate the effectiveness of the proposed method.

• Journal of the Korean Data Analysis Society
• /
• 제20권6호
• /
• pp.2853-2864
• /
• 2018

판별분석(discriminant analysis)은 새로운 개체가 입력되었을 때, 그 개체가 어느 그룹에 속하는지 예측하는데 사용되는 분석방법이다. 판별분석에서는 레이블(label)을 통해 새로운 개체를 예측하기 때문에 판별분석에서 레이블은 중요하다. 레이블 노이즈(label noise)는 관측된 레이블에 오류가 포함된 것을 의미하며, 실데이터에 발생하기 쉽고 판별성능에 영향을 미칠 수 있는 중요한 요인이다. 이를 개선하기 위해 레이블 노이즈와 레이블 노이즈에 강건한 모형들이 연구되고 있지만, 레이블 노이즈가 존재할 때 판별성능에 영향을 줄 수 있는 요인을 고려하고 이 요인들이 판별성능에 미치는 영향을 비교한 연구는 찾기 힘들다. 따라서 이 논문에서는 분류문제에서 많이 사용되는 LDA, QDA, KNN, SVM 방법을 이용하여 레이블 노이즈가 판별성능에 미치는 영향을 알아보고자 한다. 특히 판별분석의 성능과 연관이 있을 것으로 예상되는 레이블 노이즈의 발생 비율, 발생형태, 데이터의 개수에 따른 판별성능을 모의실험을 통해 살펴보았다. 그 결과, 데이터의 형태와 분석기법에 따라 레이블 노이즈가 판별성능에 영향을 미치는 정도가 다름을 확인하였다.