의상디자인학과 ; 또래문화가 프리틴 세대의 제품구매에 미치는 영향

한기향 ( Ki Hyang Han ),원명심 ( Myung Sim Won ) 건국대학교 조형연구소 2008 조형연구 Vol.16 No.-

The purpose of this study was to examine effect of peers on the purchase of preteen. Preteen is the age group between 11 to 13 who are in the high grades of elementary school and it is called tweens. They boast their intensive consumption potentials based on strong economic power and in some cases, influence the consumption of the parents. They make subjective consumptions as adults, but it is the age between youth and teenager which is a very important time to be in the peer group in developmental psychology. It is the time when they find commonness from similar behavior patterns as peer groups, especially, visible clothes. Recently, as the time to hang out with peers is increased through academic institutions after school, clothes is a good element for them to confirm their commonness with their peers since they don`t wear school uniforms as middle school and high school students. For preteen, the peer group as a conformity group has influence on the motive of purchasing products and the actual stage of purchase. The biggest characteristic of the preteen generation is the strong conformity consumption disposition. Especially, their desire to have the object when a friend has it is very strong. Preteen is good at internet does not find it difficult to find the information they want, share it with friends and desire to find and body the product of their desire. The conformity consumption condition of preteen generation is considered to be a means for relationship with friends and the big mouth influence of the leader is massive which they wish to conform with the friend they think is the leader. It is necessary to understand preteen generation cultures which are the most important time to belong to the peer group, establish marketing strategies in accordance, and educate healthy consumption cultures.

2030 세대의 아웃도어 웨어 선택속성이 구매 만족 및 재구매 의도에 미치는 영향

한기향(Han, Ki Hyang) 한국복식학회 2021 服飾 Vol.71 No.3

머신러닝을 이용한 여성 골프웨어 판매량 예측 -기상요인과 요일을 중심으로-

한기향(Han, Ki Hyang) 한국복식학회 2021 服飾 Vol.71 No.1

빅데이터를 이용한 패션인플루언서에 대한 소비자 견해 분석

한기향(Ki-Hyang Han) 한국디지털콘텐츠학회 2019 한국디지털콘텐츠학회논문지 Vol.20 No.11

Unstructured big data from social media platforms can be used as a marketing strategy to predict consumer behavior using a variety of analysis methods. This study investigated and analyzed the consumer awareness of fashion products specialized in fashion among the influencer markets that have established themselves as cell markets using big data. A frequency analysis using fashion influencer as a keyword showed that consumers are aware of text such as beauty, brands, instagram, shopping malls and marketing. The results of the N-gram analysis, “SNS-product package” and “product-interest,” show that consumers are not only interested in fashion influencers who produce products and sell on SNS markets, but are also satisfied. The result of the CONCOR analysis based on the frequency analysis and matrix analysis has been divided into three groups, “SNS market”, “influencer e-commerce”, “new influencer”, and the analysis of each group’s characteristics suggests that consumers are highly interested in fashion influencers.

패션 인플루언서의 특성이 구매의도에 미치는 영향

한기향(Ki-Hyang Han) 한국디지털콘텐츠학회 2019 한국디지털콘텐츠학회논문지 Vol.20 No.11

As influencers, which refers to the public who directly or indirectly influences others in SNS, are growing in influence, companies maximized their marketing strategies through influencer marketing. However, due to rapidly changing social environment, influencers have recently started creating a new distribution channel called influencer markets. As it is believed that this is a time when more diverse and detailed research is needed on fashion influencers who are creating new trends than preceding studies, this study tried to analyze the impact of the characteristics of fashion influencers on consumers’ willingness to buy fashion products, and used “R” which is an open source as a visualization tool. The results showed that there is no difference in perception between men and women regarding the characteristics of fashion influencers. However, men found that unlike women, the informality of fashion influencers affected their intention to purchase fashion products, indicating that there is a difference between men and women. This study aims to identify the characteristics of fashion influencers and help establish a marketing strategy suitable for target customers. It is also hoped that research using open source will be actively conducted in follow-up studies.

토픽 모델링을 활용한 아웃도어웨어 패딩 리뷰의 융복합적 내용 분석

한기향(Han, Ki Hyang) 한국전시산업융합연구원 2024 한국과학예술융합학회 Vol.42 No.2

소비자 리뷰는 온라인 쇼핑몰에서 의류를 구매한 소비자가 직접 자신의 경험이나 의견을 적은 것으로, 소비자들은 상품 구매 시 먼저 의류 상품을 구매한 소비자의 경험에 많은 의존을 하고 있다. 본 연구의 목적은 코로나-19 이후 변화된 소비자의 라이프스타일과 고프코어룩 트렌드로 인해 새로운 전성기를 맞이한 아웃도어웨어의 지속적인 성장을 위한 기초 자료를 제공하는데 있다. 연구를 위한 데이터는 국내 최대 온라인 패션 플랫폼인 ‘무신사’를 대상으로 2023년 12월 5일을 기준으로 3개월 동안 판매된 남성용 아웃도어웨어 패딩의 리뷰 1,046개를 Python을 이용하여 수집하였다. 수집된 데이터는 TEXTOM을 이용하여 형태소 분석, 감성 분석, LDA 토픽 모델링을 진행하였으며, 연구 결과는 다음과 같다. 첫째, 빈도분석결과 ‘좋다’가 615개로 가장 많이 도출되었으며, 다음으로 ‘입다(535개), 예쁘다(306개), 구매하다(273개), 같다(266개)’ 등의 순으로 나타났으며, TF-IDF는 ‘입다(489.10)’의 순위가 가장 높았으며, 다음으로 ‘좋다(471.41), 예쁘다(412.00), 같다(406.39), 구매하다(398.49)’의 순으로 나타났다. 둘째, 감성 분석을 시행한 결과, 전체 1,046개의 리뷰 중 긍정 감성이 860개(82.22%)로 가장 높게 나타났으며, 부정 감성이 186개인 17.78%로 나타났다. 셋째, LDA 토픽 모델링 결과, 총 5개의 토픽이 도출되었다. 토픽 1은 남성 아웃도어웨어 패딩을 구매한 소비자의 만족 요인으로 구성되어 있으며, 이 외에 토픽에 포함된 주요 리뷰로 선물이 있다. 토픽 1의 주제는 “소비자 만족 요인-사이즈, 가격, 배송”으로 명명하였다. 토픽 2의 주제는 “소비자 불만족 요인-사이즈, 재질, 핏(fit),”로 사이즈, 재질, 핏(fit)에 관한 소비자 불만족 요인으로 구성되어 있다. 토픽 3의 주제는 “소비자 만족 요인-색상, 디자인, 품질”로 소비자 만족 요인 중 색상, 디자인, 품질에 만족한 소비자 리뷰로 이루어져 있다. 토픽의 주제가 “리버서블”인 4번째 토픽은 리버서블 패딩에 대한 긍 · 부정의 리뷰로 이루어졌다. 5번째 토픽은 “기능성 및 활용성”을 주제로 다수의 긍정 리뷰와 약간의 부정 리뷰를 포함하고 있다. 본 연구는 온라인 패션 플랫폼에서 아웃도어웨어 패딩을 구매한 소비자가 직접 남긴 리뷰를 수집하고 분석한 연구로 최근 텍스트 마이닝 분석 방법 중 주목받고 있는 LDA 토픽 모델링을 이용하여 현재 판매되고 있는 상품에 대한 소비자의 의견을 분석하였다는 데 그 의의가 있다. Consumer reviews are written by consumers who purchase clothing from online shopping malls, and consumers rely heavily on the experience of consumers who purchase clothing products first when purchasing products. The purpose of this study is to provide basic data for the sustainable growth of outdoor wear, which has entered a new heyday due to the changed consumer lifestyle and Gorp-Core look trend after COVID-19. The data for the study was collected using Python for 1,046 reviews of mens outdoor-wear padding sold for three months as of December 5, 2023, targeting MUSINSA, the largest online fashion platform in Korea. Frequency analysis, TF-IDF, sentimental analysis, and LDA topic modeling were conducted using TEXTOM for the collected data, and the results of the study are as follows. First, the frequency analysis results were in the order of good, wear, pretty, buy, same, and TF-IDF in the order of wear, good, pretty, same, buy. Second, as a result of conducting an sentimental analysis, 860 (82.22%) out of 1,046 reviews were positive, and 186 (17.78%) were negative emotions. Third, as a result of LDA topic modeling, a total of 5 topics were derived. Topic 1 is composed of consumer satisfaction factors, and in addition, gifts are the main reviews included in the topic. The subject of Topic 1 was named Consumer Satisfaction Factor-Size, Price, and Delivery. The subject of Topic 2 is Consumer dissatisfaction factor-Size, Material, Fit, which consists of consumer dissatisfaction factors related to size, material, and fit. The subject of Topic 3 is Consumer Satisfaction Factor-Color, Design, Quality, consisting of consumer reviews that are satisfied with color, design, and quality among consumer satisfaction factors. The 4th topic, whose topic is Reversible, consisted of positive and negative reviews on reversible padding. The 5th topic included a number of positive and some negative reviews with the theme of Functionality and Utilization. This study is a study that collects and analyzes reviews left by consumers who purchase outdoor-wear padding on an online fashion platform, and is significant in that it analyzes consumer opinions on the products currently on sale by using the LDA topic modeling, which is attracting attention among text mining analysis methods.

데이터마이닝을 이용한 아웃도어웨어 판매율에 영향을 미치는 요인 분석

한기향 ( Han Ki Hyang ) 한국기초조형학회 2022 한국기초조형학회 학술발표논문집 Vol.2022 No.2

반팔 티셔츠의 온라인 리뷰 분석에 관한 융합적 분석 연구

한기향(Han, Ki hyang) 한국전시산업융합연구원 2021 한국과학예술융합학회 Vol.39 No.4

본 연구는 패션 온라인 쇼핑몰에서 반팔 티셔츠를 구매한 소비자의 리뷰를 통해, 반팔 티셔츠 구매 시 소비자가 중요하게 생각하는 요인과 긍·부정의 감정을 분석하고자 시작되었다. 본 연구의 목적은 패션 온라인 쇼핑몰에서 반팔 티셔츠를 구매한 소비자가 남긴 댓글을 분석하여 소비자의 반팔 티셔츠 구매 요인을 확인하고 소비자의 만족과 불만족에 대한 이유를 분석하는 것에 있다. 따라서 본 논문은 Python 3.7 프로그램을 이용하여 패션 온라인 쇼핑몰에 소비자가 작성한 댓글을 2021년 7월 30일을 기준으로 크롤링하였으며, 텍스톰을 이용하여 텍스트마이닝을 실시하였다. 연구 결과 및 내용은 다음과 같다. 첫째, 빈도분석과 TF-IDF, 연결 중심성 결과 ‘입은’, ‘사이즈’의 순서로 나타나, 소비자가 반팔 티셔츠 구매 시 가장 중요하게 생각하는 요인이 ‘사이즈’임을 알 수 있었다. 또한, ‘사이즈’ 이외 평가 요인으로 ‘심미적 요인’, ‘소재의 품질’, ‘세탁’, ‘활용성’인 것을 알 수 있다. 둘째, 감성 분석 결과 긍정의 비율이 부정의 비율보다 높게 나타나 구매한 반팔 티셔츠에 대해 전반적으로 만족하고 있음을 유추할 수 있었다. 긍정의 댓글은 대부분이 ‘심미적 요인’과 연관된 ‘예쁘다’, ‘귀엽다’가 많았으며, 오버핏 ‘사이즈’에 대한 긍정의 댓글도 많았다. 부정의 댓글은 상품 전체에 대한 부정의 의견보다는 상품 일부에 대해 아쉬움을 표현하는 댓글이 많았으며 ‘세탁’에 관한 염려의 글이 많이 나타났다. 이러한 연구 결과가 패션 온라인 쇼핑몰의 운영전략과 패션 브랜드의 상품기획에 도움이 될 수 있는 기초 자료로 사용되기를 바란다. 본 연구는 반팔 티셔츠에 대한 소비자의 의견을 상품이 판매되는 근접한 시점에 빅데이터 분석을 이용해 규명한 융합연구로써 그 의의가 있다. This study was initiated to analyze factors that consumers value when purchasing short-sleeved T-shirts and positive and negative emotions through reviews of consumers who purchased it at fashion online shopping malls. The purpose of this study is to analyze the comments left by consumers who purchased short-sleeved T-shirts from online fashion shopping malls, identify the factors behind consumers’ purchase of short-sleeved T-shirts, and analyze the reasons for consumer satisfaction and dissatisfaction. Therefore, this paper crawled comments written by consumers on online fashion shopping malls using Python 3.7 program as of July 30, 2021, and conducted text mining using Textom. The results and contents of the study are as follows. First, frequency analysis, TF-IDF, and degree centrality results were shown in order of ‘worn’ and ‘size’, which shows that ‘size’ is the most important factor for consumers to purchase short-sleeved T-shirts. In addition, evaluation factors other than ‘size’ are found to be ‘Aesthetic factors’, ‘Quality of Material’, ‘Washing’ and ‘utility’. Second, the sentimental analysis showed that the positive rate was higher than the negative rate, inferring that the overall satisfaction with the short-sleeved T-shirts purchased was high. Most of the positive comments were ‘pretty’ and ‘cute’ related to ‘aesthetic factors’, and there were many positive comments about overfit ‘size’. Negative comments expressed regret over some factors of the products rather than the entire product, and there were many concerns about ‘washing’. It is hoped that these findings will be used as basic materials that can help with the operation strategy of fashion online shopping malls and product planning for fashion brands. This study is meaningful as a convergence study that identifies consumers’ opinions on short-sleeved T-shirts using big data analysis at a time when products are sold.

머신러닝을 이용한 골프웨어 이미지 판별분석 정확도에 대한 연구

한기향(Ki Hyang Han),우종필(Jong Pil Yu) 한국상품학회 2022 商品學硏究 Vol.40 No.5

코로나 19 이후 골프웨어 시장이 새로운 호황기를 맞이하고 있다. 그러나, 지금은 그 어느 때보다도 골프웨어 시장의 경쟁이 심각한 상황으로, 소비자들에게 브랜드를 인지시키고 선호도를 증가시켜 매출을 활성화하기 위해서 그 무엇보다도 브랜드 아이덴티티가 중요한 시기라 사료된다. 또한, 4차 산업혁명의 다양한 기술이 패션산업에 이용되는 시점으로, 본 연구는 최근 그 이용이 증가하고 있는 머신러닝을 이용해 이미지 분석을 통한 브랜드의 차별성을 검증하는 것을 연구의 목적으로 하였다. 이에 골프웨어 시장을 주도하고 있는 ‘PXG’, ‘Titleist’, ‘NIKE golf’와 ‘ADIDAS golf’ 등 4개 브랜드의 남녀 상·하를 대상으로 머신러닝의 Neural Network, Logistic regression, Adaboost 세 가지 모델을 사용하여 이미지만으로 브랜드 구별이 가능한지를 측정하였다. 연구 결과, 머신러닝을 이용해 골프웨어 이미지만으로 브랜드 구별이 가능해 브랜드별 판별성과 유사성을 확인할 수 있었으며, 세 개의 모델 중 Neural Network( .97), Logistic Regression( .97)의 판별 정확도가 .97 이상으로 높게 나타났다. 남성 상의는 ‘NIKE golf(95%)’, 남성 하의는 ‘PXG(94%)’의 판별 정확도가 가장 높은 것으로 나타났으며, 여성 상의는 ‘PXG(95%)’, 여성 하의는 ‘NIKE golf(96%)’의 판별 정확도가 가장 높게 나타났다. 본 연구는 최근 패션산업에서 주목받고 있는 머신러닝 방법인 이미지 분석을 통해 브랜드 간 디자인의 차이를 검증하였다는 데 연구의 의의가 있다. 본 연구 결과가 학계의 패션상품 이미지 분석 연구의 초석이 되기를 바라며, 골프웨어 브랜드에서는 상품 차별화를 위한 전략에 다양한 디지털전환의 기술이 활용될 수 있는 기본 자료로 활용되기를 기대한다. Golf wear market is in a new boom after COVID-19. However, competition in the golf wear market is more serious than ever, and brand identity is considered to be more important than anything else to recognize the brand and increase preferences to boost sales. In addition, at the time when various technologies of the Fourth Industrial Revolution are used in the fashion industry, the purpose of this study is to verify the brand's differentiation through image analysis using machine learning, which has recently been increasing in use. This study used three models of machine learning: Neural Network, Logistic regression, and Adaboost to measure the brand distinction only by image for men and women's tops and bottoms of four brands, ‘PXG’, ‘Titleist’, ‘NIKE Golf’ and ‘Adidas Golf’. As a result of the study, it was possible to distinguish brands only with golf wear images using machine learning, confirming the discrimination and similarity of each brand. Among the three models, the discrimination accuracy of Neural Network ( .97) and Logistic Regression ( .97) was higher than .97. The discrimination accuracy of ‘NIKE golf (95%)’ in men's tops and ‘PXG (94%)’ in men's bottoms was the highest, while ‘PXG (95%)’ in women's tops and ‘NIKE golf (96%)’ in women's bottoms was the highest. This study is meaningful in that it verified the difference in design between brands through image analysis, a machine learning method that has recently attracted attention in the fashion industry. It is hoped that the results of this study will be the cornerstone of the academic research.

연구논문 : 초등학교 고학년 남학생들의 스포츠웨어 브랜드 충성도

한기향 ( Hyang Han Ki ),원명심 ( Myung Sim Won ) 한국의류산업학회 2013 한국의류산업학회지 Vol.15 No.1

This study examines the effectiveness of brand awareness, brand image and brand identification on brand loy-alty for the Nike brand. The subjects of this study were 336 elementary school boys. The statistical methods used for this study were factor analysis, Cornbach`s alpha analysis, confirmatory factor analysis and path analysis with SPSS 19.0 and AMOS 19.0. The results of this study were as follows. First, brand awareness had a direct effect on brand image and brand loyalty. Brand awareness also had an indirect effect on brand loyalty. Second, brand image directly or indirectly influenced brand loyalty as well as directly influenced brand identification. Third, brand identification directly influenced brand loyalty. The 5 grade group and 6th grade group had different paths; however, the most powerful path was the same as brand awareness to brand image. The results of this study will help fashion companies understand the importance of new consumer groups in their early teens or elementary school.