텍스트마이닝 분석을 통한 가상 인플루언서의 인식 및 활용 방안에 관한 연구

Abstract

The virtual Influence market is gradually expanding, and its influence is growing as it provides inspiration for tastes and lifestyles. However, research on public awareness and utilisation method regarding this phenomenon is still lacking. In this respect, this research investigated issues on virtual influencers that have appeared in the news and social media over the past two years divided into two periods based on the lifting of outdoor mask mandates during the COVID-19 pandemic. Frequency analysis, N-gram analysis, CONCOR analysis and sentiment analysis were conducted. Through the analysis, the current status of the use of virtual influencers in South Korea and societal perceptions were examined. Based on this, business insights regarding the utilisation of virtual influencers were derived. The analysis revealed that virtual influencers are expanding beyond advertising models to various fields such as finance, fashion, beauty. They are extending their roles to include singers and show hosts, creating economic ripple effects. Despite the increase in positive sentiment towards virtual influencers, there has been a significant decrease in media and social media mentions due to the increase of outdoor activities and related events after the pandemic. The appearance of virtual influencers has evolved to the point where expressions like 'cute', and 'pretty' are used, overcoming uncanny valley. It is assumed that people feel a sense of intimacy, experiencing emotions such as 'love', 'friendship', and 'expectation'. However, it is confirmed that negative perception related to technological anxiety when using new technology based products or services and skepticism about authenticity, implied by the word 'fake' still exist. This research can be a cornerstone for design research related to user journey or scenarios when planning services using virtual influences in the future by helping the public's perception and use of virtual influencers.