• 한국의류학회지
• /
• 제46권1호
• /
• pp.80-98
• /
• 2022

Artificial intelligent chatbot services have recently become common in fashion e-retailing and are expected to improve online shopping by making it easy to recommend products. This study examines whether the perceived quality of a fashion chatbot affects consumers' trust and perception of usefulness, which in turn influences satisfaction and intention to use, in accordance with the information system success model. The study also investigates differences in perceived quality and consumer response variables between high and low groups of self-efficacy. A total of 341 consumers participated in an online survey. The results revealed that information quality and system quality had a significant impact on perceived usefulness and trust, and that service quality significantly impacted trust. Perceived usefulness and trust had a positive effect on consumer satisfaction, which in turn had a positive effect on intention to use. In addition, the findings revealed that people who had higher self-efficacy showed higher scores on perceived usefulness, trust, satisfaction, and intention to use chatbots as compared to people who had lower self-efficacy. This study suggested theoretical implications by applying the information system success model theory to fashion chatbot studies. It also suggested practical implications for e-commerce marketers developing retail strategies.

• 패션비즈니스
• /
• 제27권3호
• /
• pp.132-149
• /
• 2023

This study aims to analyze Instagram hashtags based on big data to investigate changes in consumer trends and perceptions of vegan fashion, and to derive strategies for revitalizing vegan fashion brands based on derived results. Among social media, Instagram was selected as a collection channel, and Instagram hashtags for 'Vegan Fashion' were collected from July 1, 2021 to December 31, 2021. After conducting semantic network analysis with the Ucinet 6 program based on the collected data, the CONCOR analysis on vegan fashion showed the following four clusters: 'Veganism practiced with fashion', 'Bag type of vegan fashion brand', 'Sharing vegan fashion', and 'Diversification of eco-friendly products'. Analysis results showed that the Instagram hashtag for vegan fashion confirmed the MZ generation's increased interest in vegan fashion and their thoughts to recommend and share frequently used items or brand products to people around them. CONCOR analysis of vegan fashion brands showed the following four groups: 'Differentiating the material of vegan bags', 'Eco-friendly products of vegan fashion brands', 'Interest in vegan shoes', and 'Donation campaign of vegan fashion brands'. CONCOR analysis on Meaningout showed the following four clusters: 'MZ Generation's Meaningout Start-up', 'Recommendation Platform for Skin Products', 'Value Consumption Trend for Eco-friendly Clothing', and 'Interest in Eco-friendly Packaging'. The results of this study on vegan fashion, a practical eco-friendly movement that can require changes in social responsibility and perception as issues that directly affect animals, the environment, and humans, are expected to provide basic data to help domestic vegan fashion brands develop marketing strategies.

• 복식문화연구
• /
• 제14권3호
• /
• pp.483-496
• /
• 2006

The purpose of this study was to develop a pragmatic fashion internship program as a following study of 'An analysis of the consciousness of the professors and college students for the fashion industry internship'. The methods of the quantitative analysis and some case studies were used for this study. The results of this study were as follows : Internship program has 6 steps ; ready, introduction, selection, operation, reputation, management. It will help professors and college students to execute the program systematically. The contents of the fashion internship program model for professors were as follows: 8 recruiting methods to find internship companies, the contents of the internship agreement, 6 types of internship, the contents of the recommendation letter, advanced education contents for a fashion internship, e-mentoring program model, a curriculum of the cultural study and the major course, evaluation standard elements of the intern students, the methods of evaluation etc.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2018년도 춘계학술발표대회
• /
• pp.500-503
• /
• 2018

최근 스마트폰의 보급률이 높아지면서 인터넷 쇼핑몰의 접근성이 용이해지고 있고 그로 인해 사용자들의 인터넷 쇼핑의 이용이 보편적이게 되었다. 그 중 여성 의류 분야는 많은 비중을 차지하고 있으며 현재도 꾸준히 성장하고 있는 추세이다. 많은 여성 소비자들은 개인의 취향에 맞는 의류들을 추천받기를 원한다. 본 논문에서는 협업 필터링에서 발생하는 cold start 문제를 이름, 나이, 선호 스타일, 자주 사용하는 쇼핑몰 등 개인 성향을 이용하여 해결하는 협업 필터링과 개인 성향을 이용한 여성 의류 쇼핑몰 추천 시스템을 제안한다.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2020년도 추계학술발표대회
• /
• pp.37-40
• /
• 2020

날씨에 대한 인류의 관심은 인류 역사가 시작되면서 지금까지 예측하며 관심 영역인 만큼 인류에게 끼치는 영향이 크다. 초기 인류에게 있어서 의류는 생존을 위한 생존 도구에서 현재는 패션의 영역으로 자기를 표출하거나 자신에게 가장 어울리는 옷을 찾기 위한 욕구로 발전해 왔다. 따라서 본 논문에서는 날씨에 따른 개인의 체감온도와 해당 날씨에 가장 선호하는 의상을 분석하고, 예측하며 추천해주는 시스템을 제안한다. 제안하는 시스템은 지속적인 유지 관리를 통해 보완해 나간다면 날씨와 패션 분야에서 다양한 접목을 하는 등 기술발전을 할 것으로 기대된다.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2020년도 춘계학술발표대회
• /
• pp.393-396
• /
• 2020

추천시스템은 대량의 정보를 이용하여 특정 사용자가 선호할만한 상품의 리스트를 추천하는 것이다. 현재 추천시스템으로 유명한 Netflix, Amazon, Youtube 등은 기업내의 상품 및 사용자 데이터를 토대로 이루어 졌으나 스타트 업 및 소규모 기업이 추천 시스템을 구축하기 위해선 기반이 될 데이터셋 자체가 없으며 데이터 수집에도 한계가 있다. 본 논문에서는 옷 추천 시스템 구축을 위해 특정 기업만이 아닌 모든 의류매장들이 사용할 수 있는 데이터 셋 구축 방법에 대해 제안하며, 고객 데이터 셋 구축을 위한 텍스트 데이터 마이닝 처리 과정과 결과에 대해 기술한다.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2022년도 추계학술발표대회
• /
• pp.643-645
• /
• 2022

많은 온라인 쇼핑몰에서 색상 기반 필터링 서비스나 추천 시스템을 제공하지만, 수동 분류는 많은 시간이 들고 오류 위험이 있다. 본 연구의 실험에서는 먼저 분석할 의류 이미지를 실루엣 분석으로 수행한 경우와 수행하지 않는 경우의 k-평균 군집화 알고리즘으로 가장 우세한 색상 군집의 중심값을 도출하는데, 만약 군집 개수가 2개 이상이면 보다 큰 군집의 중심값만을 고려한다. 이 중심값을 이용해 사전 학습한 k-최근접 이웃 알고리즘으로 색상 클래스를 분류한다. 실험 결과 실루엣 분석을 수행하지 않은 k-평균 군집화 알고리즘을 사용한 분류 방식이 정확도와 수행 시간 모두 매우 준수하였으나, 배경색이 존재하여 의류 색 분석에 영향을 줄 수 있는 경우 잘못 분류한다는 문제도 있다.

• 한국의류학회지
• /
• 제42권4호
• /
• pp.671-688
• /
• 2018

This study investigates the characteristics of VR stores that emerged as new fashion communication media. Two case studies on Hyundai and eBay VR Department stores were conducted along with a discussion of the function and meaning of the fashion VR store. The results showed that both stores provide novel shopping experiences; however, the two were differentiated in terms of production method and technology implementation level. Functional aspects such as providing shopping efficiency and purchasing service was insufficient in both stores. Instead, they were complementing by means of product rotation, recommendation system, voice guidance, or linkage with an online shopping mall. In experiential aspects, both stores provided a strong sense of immersion. Hyundai VR store enhanced immersion with a high resolution image of a real offline store; however, it lacked in the ability to provide multisensory stimulation such as kinetic sense or auditory stimulation. The eBay VR Department store intensified the immersion experience by providing auditory stimulation as well as visual stimulation that enhanced the speed and distance sense through the utilization of animation. However, the extent of experience was limited in terms of agency and transformation because of the low interactivity found in both store systems.

• 한국산학기술학회논문지
• /
• 제11권12호
• /
• pp.5113-5119
• /
• 2010

인체측정학(Anthropometry)은 자동차, 가전제품, 의료기기 및 스포츠 용품 등 다양한 분야에서 소비자의 만족도와 사용 효율성을 최적화하기 위해 조사되고 연구되어 왔다. 하지만 아직까지 인체 측정 방식은 계측자를 이용한 직접측정 또는 스캐너나 디지털 측정기 등 고가의 장치에 의존적인 방법에서 벗어나지 못하고 있다. 따라서 본 연구는 사진으로부터 신체인식 알고리즘을 이용하여 신체사이즈를 자동으로 추출하고, 의류추천 및 자전거 피팅사이즈 서비스 등에 활용가능성을 제시한다. 이를 위해 Haar-like features와 AdaBoost 알고리즘을 이용하여 빠른 속도와 높은 정확도로 영상콘텐츠 내에서 신체 영역을 검출한다. 이후 AAM(Active Appearance Model)을 이용하여 특징점을 검출하고 도출된 측정치에 최적화된 상품을 추천하는 지능형 모듈 시스템을 구현하고 성능평가를 제시한다.

• 복식
• /
• 제60권6호
• /
• pp.89-100
• /
• 2010

This study is intended to analyze current status of costume in the cultural industry and explore feasibility of integrating costume as important cultural contents in the digital era. Among the websites for contents as the cultural archetypes, some have strong foundations while others don't. First, as for various archetypal characters restored based on the Korean history, most of their clothes focus on shapes and colors but lack details. They should be produced in 3D to provide back views or textile patterns to users. To ease understanding of the history of costume, user-friendly services such as launch of a pop-up window when users click on specific contents in question should be available for detailed information. At least there should be a link to other related sites where users can conveniently find more details. Second, some sites have too much data under one subject, increasing complexity and undermining orderliness. As a result, it takes long time to identify the site map. In this case, it is required to rearrange the contents with Quick View by subject and related links for in-depth study. Third, each subject is important to develop the archetypes for a variety of purposes. Creation of design derived from them or their commercialization can be an example but these activities should not restrict imagination of users or degrade the value of the archetypes. Last, it is needed to adopt validation system to detect the needs for a regular update (renewal) and to fix errors. We found many servers whose operation is not stable in general. When these technical issues are addressed for stable operation, users will rely on the sites to utilize them for their purpose of developing the cultural archetypes. In conclusion, advancement of www.culturecontent.com is essential. Based on efficient management and operation of the system, the quality of contents would be increased and multi-faceted advertising campaigns focusing on needs should be launched, to promote application of the contents. This is the recommendation for the future of the cultural archetype industry in Korea.