• Smart Media Journal
• /
• v.13 no.4
• /
• pp.57-64
• /
• 2024

Attributes such as material, color, and fit in fashion images are important factors for consumers to purchase clothing. However, the process of classifying clothing attributes requires a large amount of manpower and is inconsistent because it relies on the subjective judgment of human operators. To alleviate this problem, there is a need for research that utilizes artificial intelligence to classify clothing attributes in fashion images. Previous studies have mainly focused on classifying clothing attributes for either tops or bottoms, so there is a limitation that the attributes of both tops and bottoms cannot be identified simultaneously in the case of full-body fashion images. In this study, we propose a deep learning model that can distinguish between tops and bottoms in fashion images and classify the category of each item and the attributes of the clothing material. The deep learning models ResNet and EfficientNet were used in this study, and the dataset used for training was 1,002,718 fashion images and 125 labels including clothing categories and material properties. Based on the weighted F1-Score, ResNet is 0.800 and EfficientNet is 0.781, with ResNet showing better performance.

• Journal of Intelligence and Information Systems
• /
• v.26 no.4
• /
• pp.67-86
• /
• 2020

Recommender Systems have been huge influence users and business more and more. Recently the importance of E-commerce has been reached rapid growth greatly in world-wide COVID-19 pandemic. Recommender system is the center of E-commerce lively. Top ranked E-commerce managers mentioned that recommender systems have a major influence on customer's purchase such as about 50% of Netflix, Amazon sales from their recommender systems. Most algorithms have been focused on improving accuracy of recommender system regardless of novelty, diversity, serendipity etc. Recommender systems with only high accuracy cannot satisfy business long-term profit because of generating sales polarization. In addition, customers do not experience enjoyment of shopping from only focusing accuracy recommender system because customer's preference is changed constantly. Therefore, recommender systems with various values need to be developed for user's high satisfaction. Reranking is the most useful methodology to realize diversity of recommender system. In this paper, diversity of recommender system is represented through constructing high similarity with users who have different preference using each user's purchased item's category algorithm. It is distinguished from past research approach which is changing the algorithm of recommender system without user's diversity preference level. We tried to discover user's diversity preference level and observed the results how the effect was different according to user's diversity preference level. In addition, graph-based recommender system was used to show diversity through user's network, not collaborative filtering. In this paper, Amazon Grocery and Gourmet Food data was used because the low-involvement product, such as habitual product, foods, low-priced goods etc., had high probability to show customer's diversity. First, a bipartite graph with users and items simultaneously is constructed to make graph-based recommender system. However, each users and items unipartite graph also need to be established to show diversity of recommender system. The weight of each unipartite graph has played crucial role changing Jaccard Distance of item's category. We can observe two important results from the user's unipartite network. First, the user's diversity preference level is observed from the network and second, dissimilar users can be discovered in the user's network. Through the research process, diversity of recommender system is presented highly with small accuracy loss and optimalization for higher accuracy is possible controlling diversity ratio. This paper has three important theoretical points. First, this research expands recommender system research for user's satisfaction with various values. Second, the graph-based recommender system is developed newly. Third, the evaluation indicator of diversity is made for diversity. In addition, recommender systems are useful for corporate profit practically and this paper has contribution on business closely. Above all, business long-term profit can be improved using recommender system with diversity and the recommender system can provide right service according to user's diversity level. Lastly, the corporate selling low-involvement products have great effect based on the results.

• Journal of Intelligence and Information Systems
• /
• v.26 no.1
• /
• pp.97-117
• /
• 2020

Product evaluation criteria is an indicator describing attributes or values of products, which enable users or manufacturers measure and understand the products. When companies analyze their products or compare them with competitors, appropriate criteria must be selected for objective evaluation. The criteria should show the features of products that consumers considered when they purchased, used and evaluated the products. However, current evaluation criteria do not reflect different consumers' opinion from product to product. Previous studies tried to used online reviews from e-commerce sites that reflect consumer opinions to extract the features and topics of products and use them as evaluation criteria. However, there is still a limit that they produce irrelevant criteria to products due to extracted or improper words are not refined. To overcome this limitation, this research suggests LDA-k-NN model which extracts possible criteria words from online reviews by using LDA and refines them with k-nearest neighbor. Proposed approach starts with preparation phase, which is constructed with 6 steps. At first, it collects review data from e-commerce websites. Most e-commerce websites classify their selling items by high-level, middle-level, and low-level categories. Review data for preparation phase are gathered from each middle-level category and collapsed later, which is to present single high-level category. Next, nouns, adjectives, adverbs, and verbs are extracted from reviews by getting part of speech information using morpheme analysis module. After preprocessing, words per each topic from review are shown with LDA and only nouns in topic words are chosen as potential words for criteria. Then, words are tagged based on possibility of criteria for each middle-level category. Next, every tagged word is vectorized by pre-trained word embedding model. Finally, k-nearest neighbor case-based approach is used to classify each word with tags. After setting up preparation phase, criteria extraction phase is conducted with low-level categories. This phase starts with crawling reviews in the corresponding low-level category. Same preprocessing as preparation phase is conducted using morpheme analysis module and LDA. Possible criteria words are extracted by getting nouns from the data and vectorized by pre-trained word embedding model. Finally, evaluation criteria are extracted by refining possible criteria words using k-nearest neighbor approach and reference proportion of each word in the words set. To evaluate the performance of the proposed model, an experiment was conducted with review on '11st', one of the biggest e-commerce companies in Korea. Review data were from 'Electronics/Digital' section, one of high-level categories in 11st. For performance evaluation of suggested model, three other models were used for comparing with the suggested model; actual criteria of 11st, a model that extracts nouns by morpheme analysis module and refines them according to word frequency, and a model that extracts nouns from LDA topics and refines them by word frequency. The performance evaluation was set to predict evaluation criteria of 10 low-level categories with the suggested model and 3 models above. Criteria words extracted from each model were combined into a single words set and it was used for survey questionnaires. In the survey, respondents chose every item they consider as appropriate criteria for each category. Each model got its score when chosen words were extracted from that model. The suggested model had higher scores than other models in 8 out of 10 low-level categories. By conducting paired t-tests on scores of each model, we confirmed that the suggested model shows better performance in 26 tests out of 30. In addition, the suggested model was the best model in terms of accuracy. This research proposes evaluation criteria extracting method that combines topic extraction using LDA and refinement with k-nearest neighbor approach. This method overcomes the limits of previous dictionary-based models and frequency-based refinement models. This study can contribute to improve review analysis for deriving business insights in e-commerce market.

• Journal of Korean Academy of Nursing
• /
• v.28 no.2
• /
• pp.441-456
• /
• 1998

The purpose of this study was to develop a reliable and valid instrument to measure hope for cancer patients in Korea. This Hope Scale(Kim & Lee Hope Scale ; KLHS ) was developed based on not only critical universal attributes explaining both basic hope (generalized hope) and specific hope but also particular characteristics varing from culture and situation, which were revealed in a comprehensive review of the literature. Initially 60 items were generated from three sources : 36 items from the Q-sample used in the Kim's study, 1992, 21 representative items(statements) from the rest Q-population of the above study, 3 items related to the newly discovered category in the new qualitative study using 10 open ended question(death and dying) from the new qualitative study on the 20 cancer patients. At first 3 items were eliminated by the critique of the content validity experts, who were high experienced nurse, nursing professors. And then 4 items were eliminated in consideration of corrected item total correlation coefficiency, theoretical framework of this study. After that, 14 items were eliminated in comparing two or three items identified with the same meaning in each factor by this research team with factor loading and communality. This Hope Scale was finally constructed with 39 items. Psychometric evaluation was done on 492 adults(104 cancer patients, 388 adults who imagined who were cancer patients ranging from 18 to 76 years old. The results revealed high internal consistency Alpha coefficiency of .9351. Princial Component Factor Analysis with Varimax Rotation resulted in 8 factors with more than 1.0 of Eigenvalue. Referring to Eigenvalues, percent of variances(>60%), reproduced correlation matrix, and our theoretical framework, we decided the eight factors were the best1 solution to represent hope dimensions sufficiently. The eight factors were "confidence in possibility of cure", "sense of internal satisfaction", "being in communion", "meaning of life", "Korean hope perspectives", "belief in god", "self confidence", "self-worth". Among these factors, "confidence in possibility of cure", "sense of internal satisfaction", "Korean hope perspectives" were identified as different hope dimensions from those of Nowotny Hope Scale and Herth Hope Scale. There was significant negative correlation of r=-.4736 between this hope scale and Beck Hopelessness Scale (BHS), and significant positive correlation of r=.3685 between this hope scale and Life Orientation Test (LOT) which indicate convergent and discriminant validity. The range of hope scores was from 71 to 244, with a mean of 171.97(SD=28.16).

• The Journal of the Korea Contents Association
• /
• v.20 no.4
• /
• pp.571-581
• /
• 2020

There are large, medium, and small items in the evaluation test for medical interpreter professionals personnel and the criteria and level are not clear, which may cause confusion for those preparing for the test. Therefore, we would like to suggest that the qualification requirements for the medical translation ability test and the criteria for the evaluation items are consistent with the medical system in Korea. The survey on the medical interpreter competency test conducted was collected from domestic and foreign data, compared with similar test and overseas test. We also examined the perception of the test by experts with experience in developing and interpreting the medical interpretation test. As a result, in the 'International Culture' evaluation category, 'Language-oriented culture' was changed to 'Medical-oriented culture' and 'Interpretation ethics' was changed to 'Medical interpretation ethics'. In the evaluation items of the hospital system, the 'Medical Dispute Mediation Act', which is 「ACT ON MEDIATION OF MEDICAL DISPUTES」 of the middle item was changed the 「ACT ON REMEDIES FOR INJURIES FROM MEDICAL MALPRACTICE AND MEDIATION OF MEDICAL DISPUTES」 and the Act also reduced the four items related to the 'Medical Tourism Law' to two and added the 「ACT ON SUPPORTING THE ADVANCEMENT OF MEDICAL OVERSEAS AND ATTRACTING FOREIGN PATIENTS」. If the Medical Interpretation Proficiency Test is prepared in accordance with the medical culture of Korea, it is expected that there will be a stable opportunity for professionals who pass the examination to act as experts.

• Journal of the Korean Society of Food Culture
• /
• v.20 no.1
• /
• pp.1-14
• /
• 2005

This study was conducted to develop and evaluate food menus after investigating eating behaviors and food preferences of preschool children. Parents of the preschool children In 2 Gumi City day-care centers completed questionnaires in July 2003, which were used to assess eating behaviors and food preferences of their preschool children. Analysis of the questionnaires led to the development of the menus. Menus (n=10) were developed for five different categories (two menus were developed for each category) including general menu, unbalanced menu, anemia menu, obese menu, and traditional menu. The preschool children(n=656) evaluated the menus as they were provided with each them throughout October 2003. The five score scale method was used to evaluate taste, smell, looks, texture and general preference of each menu. Results in eating behaviors showed that 70.7% of preschool children had unbalanced eating behaviors. No gender based differences in eating behaviors were found, but in regard to food preferences boys tended to prefer carbonated drinks more than girls. Results indicated that among all menus, fruit ranked highest $(3.97{\pm}0.65)$ for food preference, and vegetables ranked lowest for food preference $(2.46{\pm}0.68)$. Food preference in regard to cooking process indicated the highest preference was for fried foods $(3.80{\pm}0.68)$ and the lowest preference was for raw vegetables $(2.61{\pm}1.27)$ and namul $(2.85{\pm}1.13)$. Preference for taste ranked the highest $(4.30{\pm}0.91)$ but preference for looks recorded the lowest $(3.95{\pm}0.89)$. Of all the foods in the menus, steamed tofu rated the highest for individual food item preference, while tuna sesame leaf rice rated the lowest preference. Statistical analysis of interrelationships among food taste, smell, looks, texture and general preference were significant (p<0.0l). Results from this study suggest that various factors including food taste, smell, looks, and texture influence the food preferences of preschool children. Therefore, it is concluded that by developing a variety of appetizing menus for use at home and in day-care centers, containing varied food items and cooking methods, preschool children will be encouraged to increase their food preferences and to establish appropriate eating behaviors.

• The Journal of the Korea Contents Association
• /
• v.15 no.8
• /
• pp.588-597
• /
• 2015

This study is to distribute 5-year-old nuri curriculum science education contents in child picture books, and to investigate the categorical difference of science education contents between domestic and foreign picture books and among genres. The subjects were 219 picture books for children from 4 to 7, listed in which is published by Children's Book Study Group in 2012 and 2013. The research tool was from the article of 5-year-old nuri curriculum nature study, to analyze the contents of young children science education in the child picture books. Content analysis categories was set to two upper-categories and seven sub-categories. Research data were calculated in the analysis of the frequency and percentage of each item's category by the method of analysis conformity. In conclusion, first, in the analyzed result of the upper categories of young children science education contents in 219 picture books, the frequency appeared in order of 'Curious to maintain and expand', 'Learn living things and the natural environment', 'To explore the investigation technique', 'To enjoy the investigation technique', 'Utilize simple tools and machines', 'To search objects and materials', 'Learn natural phenomena'. Second, in the compared result between the domestic and foreign picture books and among genres, "scientific inquiry" is appeared more than "fostering an attitude of exploration".

• Journal of Korean Academy of Nursing Administration
• /
• v.3 no.1
• /
• pp.37-50
• /
• 1997

The purpose of this study was to identify the gap between need-level and demonstration-level in nursing managerial competencies. In addition, the study proposes solutions to narrow this gap. The results of this study are as follows : 1) The mean score for need-level of each item was 4.0, and for demonstration-level, 3.5. This indicates that military hospital head nurses demonstrate a higher level of managerial competencies than the moderate level on all items. But items which were related to resource/ cost/ information managament, staff development management and professionalism management got relatively low ratings in the need-level. 2) The mean score for need-level of each category was 4.14, and for demonstration-level, 3.53. Categories on the individual dimension got a higher rating than categories on the group or organization dimension in both need-level and demonstration level. 3) The gap between need-level and demonstration-level appeared in all items(p<.05) and categories(p<.001). Although the gap was relatively low, it indicates that it is essential to plan a developmemt program for all nursing management competencies for military hospital head nurses. 4) There were significant differences in the gap between need-level and demonstration-level according to specific characteristics of the subjects. The gap did not appear in many categories on the individual dimension where the number of nursing staff was more than 10, a major grade, ICU head nurse or for head nurses having a long career. 5) Need-level and demonstration-level showed a difference according to specific characteristics of the subjects, because need-level and demonstration-level were higher where the number of nursing staff was more than 10, a major grade, and for ICU or Medical ward head nurses. The categories which showed need-level difference and demonstration-level differences according to specific characteristics of the subjects existed almostly completely in the group and organization dimension. Gap-level differences according to the number of hospital bed existed in only two categories. 6) The general causes of the gap were indicated to be 'Knowledge/ skill/ experience deficit', 'Limitation of rules and systems/ Inappropriate organizational environment' for most items, categories, and dimensions. The results of this study indicate that extensive competency developing strategies must be developed, because a gap was found in all items and categories. Specially, there is a need to concentrate attention on competencies in the group and organizational dimension which had a wider gap level. And it is important to take an individual approach according to the cause for each gap. In order to produce effective competency developing strategies, top managers must have sinsights into the importance of nursing staff development and nursing officer's efforts to develop themselves must be achieved. Further multi-dimensional(education, personnel-policy, nursing/ organizational environment) solutions to the gap must be developed and utilized.

• KDI Journal of Economic Policy
• /
• v.34 no.1
• /
• pp.135-168
• /
• 2012

This study analyzes ways to detect protest responses (hereafter, PR zero-bid) in the contingent valuation method (CVM). In order to distinguish PR zero-bids from true zero-bids (non-PR zero bids), this study adopts the concept of the implicit willingness to pay employing the Hicksian compensating surplus and the Taylor's 1st order approximation. When a respondent proposes a zero-bid (i.e., WTP=0) and chooses a PR filtering item to indicate that her implicit WTP is not necessary zero, her response is identified as a PR zero bid. PR filtering items falling into the PR zero bids category include the uncertainty of information, distrust in the government and project achievement, disagreement to project plans, discontent with the fairness of public works and their payment method and animosity against the CVM itself. The empirical analysis shows that PR zero bids take place systematically in particular respondent groups: respondents who have never used similar facilities before nor plans to use the facility provided by the public project, the employed, and low income groups. In conclusion, the study suggests that a CVM questionnaire needs to be designed carefully to minimize problems associated with PR zero bids and the potential risks of having sample selection bias should be concerned.

• Journal of the Korean Solar Energy Society
• /
• v.34 no.2
• /
• pp.32-43
• /
• 2014

This study set out to identify the importance of each factor influencing facility selection with a survey among public medical facilities under the category of public buildings and apply the importance of economy, technology and environment with the weighting factor method, thus proposing optimal application plans. The research content of each section can be summarized as follows:1) Estimated energy consumption according to the energy simulation was 65,129MWh/yr, which was 18.7% higher than that according to the calculation equation. Of the energy consumption, more than 80% was used by heating and cooling facilities and construction facilities, and 20% was used by electronics such as medical equipments and in and outdoor lighting. 2) The results of a survey on the factors influencing the importance when selecting a new and renewable energy system reveal that the upper items had a priority in economy, environment, and technology in the descending order and that the lower item shad a priority in initial investments, maintenance and repair costs=energy costs, supply reliability, energy efficiency and $CO_2$ emissions in the descending order. 3) The application alternatives were analyzed in economy, technology, and environment. As a result, a geothermal system turned out to be the most excellent one a cross all the upper and lower comparison items. Of the other systems, a solar thermal system was superior in initial investments, maintenance and repair costs, and energy efficiency, where as a photovoltaic system was superior in energy costs, supply reliability, and $CO_2$ emissions. 4) As for the mixed application ratio among economy, technology, and environment, when the percentage of a geothermal system was approximately 80% or higher in anew and renewable energy system, it was the best and most optimal application plan.