• Journal of Information Technology Applications and Management
• /
• v.29 no.3
• /
• pp.43-55
• /
• 2022

In this study, with the goal of developing a deep learning-based product recommendation model for effective matching of influencers and products, a deep learning model with a collaborative filtering model combined with generalized matrix decomposition(GMF), a collaborative filtering model based on multi-layer perceptron (MLP), and neural collaborative filtering and generalized matrix Factorization (NeuMF), a hybrid model combining GMP and MLP was developed and tested. In particular, we utilize one-class problem free boosting (OCF-B) method to solve the one-class problem that occurs when training is performed only on positive cases using implicit feedback in the deep learning-based collaborative filtering recommendation model. In relation to model selection based on overall experimental results, the MLP model showed highest performance with weighted average precision, weighted average recall, and f1 score were 0.85 in the model (n=3,000, term=15). This study is meaningful in practice as it attempted to commercialize a deep learning-based recommendation system where influencer's promotion data is being accumulated, pactical personalized recommendation service is not yet commercially applied yet.

• The Journal of Information Systems
• /
• v.32 no.4
• /
• pp.163-188
• /
• 2023

Purpose The study examines the impact of the level of anthropomorphism (both in appearance and behavior) of virtual humans on user liking. It investigates whether this relationship is mediated by social comparison experiences, with the moderated mediation effect of users' desire for self-improvement. Design/methodology/approach A between-groups experimental design was employed to examine the impact of different levels of appearance(low/mid/high) and behavior(low/high) anthropomorphism on user liking of virtual humans. The experiment was conducted in an online environment, and participants were randomly exposed to one of six stimuli, which were Instagram-like posts. Findings The results indicate that as virtual humans become more anthropomorphic, they have a positive impact on user liking. However, once the level of anthropomorphism in appearance reaches a certain point (mid vs high), there is no significant difference in user liking. Users who perceive virtual humans as highly anthropomorphic tend to engage in more social comparison experiences, which positively affects their liking for these virtual humans. Conversely, individuals with a high desire for self-improvement found that the positive effect of appearance anthropomorphism on liking through social comparison experiences was reduced. The study extends the application of social comparison theory by examining its impact on influencer marketing with virtual beings. It provides valuable insights for the formulation of influencer marketing strategies using virtual humans.

• Journal of the Korean Society of Clothing and Textiles
• /
• v.48 no.2
• /
• pp.282-299
• /
• 2024

This study aimed to examine the impact of virtual influencers' characteristics on purchase intentions through attitudes toward the influencer and the brand, and to identify which factors are important depending on involvement. We used a scenario-based online survey of 320 female consumers in their twenties and thirties, and analyzed their responses through structural equation modeling using AMOS 21.0. Virtual influencers' attractiveness, reliability, familiarity, and virtuality had significant effects on consumers' attitudes toward those influencers, while their attractiveness, familiarity, and virtuality had significant effects on consumers' brand attitudes. Notably, in contrast to the other variables, virtuality had a negative effect. In addition, consumers' attitudes toward the virtual influencer significantly affected their brand attitudes and purchase intentions. We also analyzed which characteristics had significant impacts on high- and low-involvement groups. We found that reliability had the greatest influence on purchase intentions in the high-involvement group and that familiarity had the greatest influence on purchase intentions in the low-involvement group, which confirmed that the variables affecting purchase intentions differ depending on the level of involvement.

• Journal of Service Research and Studies
• /
• v.13 no.1
• /
• pp.26-47
• /
• 2023

In the age of digital media transformation, the rapid rise of social media has changed the paradigm of traditional marketing techniques by leveraging the influence of influencers. However, the influence of influencers cannot be freed from ethical issues that arise as individuals, so virtual influencers are emerging as a countermeasure. This study is a study on how to increase the influencer effect of virtual models with a focus on the MZ generation in medical service. This study investigated whether respondents in their 40s or younger were aware of 'Rosy', a virtual influencer, and then conducted a survey on those who recognized 'Rosy'. As a result of this study, first, both cognitive and emotional motivation had a positive influence on fanship and attractiveness for virtual influencer. In addition, it was found that there was a difference in follow motive according to gender. Second, in order to lead to the intention of visiting hospitals, which is the medical service industry, only the cognitive motives with useful and reliable information and useful information for the virtual influencer were found to be significant in intention to visit.

• Information Systems Review
• /
• v.21 no.4
• /
• pp.197-220
• /
• 2019

This study aims to identify consumer motivations of open innovation project participation from digital transformation perspective. By extending a traditional intrinsic/extrinsic motivation framework, we propose a three-dimensional perspective of the self-driven, firm-driven, and sociality-driven motivations. This reveals the significance of the social effects of open innovation projects as an example of digital transformation by categorizing the motivations based on the 'influencer' of the motivation building and by highlighting the importance of sociality as an influencer. As a result, self-efficacy is identified as a key motivation when the influencer exists internally. Economic incentive and firm reputation are identified when the influencer exists externally. Finally, competition, peer evaluation and social contributions are identified when the influencer exists socially. The role of knowledge type sought through innovation projects is further introduced to explain its moderating effects on motivations. The study is validated in two steps. First, we investigate four cases of open innovation projects and examine what motivations are highlighted in each context. Second, we collect survey data from 203 online game users and ask them on their motivations. The results confirm most of our hypotheses and highlight the significance of sociality in the knowledge-seeking process in open innovation projects. This study largely contributes to digital transformation literature by extending the view of motivation and examining the moderating role of knowledge involved in the projects.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2018.01a
• /
• pp.207-210
• /
• 2018

인플루언서(Influencer)란 소셜 미디어에서 유난히 많은 영향력과 파급효과를 가지고 오는 사람들을 말하며, 이들이 만들어내는 콘텐츠는 이제는 자신들의 브랜딩을 넘어선 커머스(Commerce) 효과를 발휘하고 있다. 본 논문에서는 소셜 웹 그리고 공공데이터를 중심으로 뷰티 빅데이터와 방송 콘텐츠 빅데이터를 수집하고 분석하여 상호 상관성에 기반하여 화장품 관련 기업에서 CRM(Customer Relation Management), PLM(Product Lifecycle Management, SCM(Supply Chain Management System) 등의 경영정보시스템과 연계한 뷰티 분야에 최적화된 통합 경영정보시스템을 제안한다.

• International Journal of Advanced Culture Technology
• /
• v.10 no.1
• /
• pp.47-54
• /
• 2022

We analyze and compare the structure of the networks of Netflix, Disney+, and OCN, which are forerunners in OTT market, on Twitter. This study employs NodeXL pro as a visualization software package for social network analysis. As a result of the comparison with values of Vertices, Connected Components, Average Geodesic Distance, Average Betweenness Centrality, and Average Closeness Centrality. Netflix has comparative advantages at Vertices, Connected Components, and Average Closeness Centrality, OCN at Average Geodesic Distance, and Disney+ at Average Betweenness Centrality. Netflix has a more appropriate social network for influencer marketing than Disney+ and OCN. Based on the analysis results, the purpose of this study is to explain the structural differences in the social networks of Netflix, Disney+, and OCN in terms of influencer marketing.

• Asia Marketing Journal
• /
• v.24 no.2
• /
• pp.39-50
• /
• 2022

Live streaming has become one of the most important communication tools for influencers to synchronously interact with viewers. It is critical to understand the effect of the reciprocal and synchronously interactive relations built between influencers and viewers, so-called trans-parasocial relations, in the context of live streaming. In this study, we investigate how trans-parasocial relations impact viewers' stickiness and purchase intention on live streaming platforms. Furthermore, we investigate fanship as a mediating factor in the relationship between trans-parasocial relations and viewers' behaviors. Overall, the results reveal significant direct and indirect effects of trans-parasocial relations on viewers' stickiness and purchase intention. Higher trans-parasocial relations further lead to stronger viewers' fanship toward influencers and increases their willingness to stay longer or make purchases on live streaming platforms. These findings further the understanding of influencer-viewer relations and viewers' behavior on live streaming platforms and provide valuable insights into influencer marketing and live streaming.

• International Journal of Internet, Broadcasting and Communication
• /
• v.15 no.2
• /
• pp.31-38
• /
• 2023

This paper introduces a real-time face swapping system that enables video influencers to swap their faces with arbitrary generated face images of their choice. The system is implemented as a Django-based server that uses a REST request to communicate with the generative model,specifically the pretrained stable diffusion model. Once generated, the generated image is displayed on the front page so that the influencer can decide whether to use the generated face or not, by clicking on the accept button on the front page. If they choose to use it, both their face and the generated face are sent to the landmark extraction module to extract the landmarks, which are then used to swap the faces. To minimize the fluctuation of landmarks over time that can cause instability or jitter in the output, a temporal filtering step is added. Furthermore, to increase the processing speed the system works on a reduced set of the extracted landmarks.

• Review of Korea Contents Association
• /
• v.17 no.4
• /
• pp.14-18
• /
• 2019