Summary: | 碩士 === 國立中興大學 === 行銷學系所 === 107 === Social media has become one of the major channels to communicate with young people. Several brands invite Youtuber as the celebrity to promote their product via online video platform. Those who view online videos or ads could write their thoughts immediately, in turn, lead to accumulate a lot of text data. Most of Youtuber like to use humour appeal when they promote the brand; however, humour is not always the case in persuading the audiences. Thus, this research considers a boundary condition (i.e., brand concept) to explore the relationship between humour appeal, emotion, and purchase intention through Sentiment Analysis.
The top ten popular Youtuber were identified from a known internet platform. To eliminate the messages of the film are less than 1000, six films were selected from a pretest due to the differential in humour appeal and brand concepts. This study conducted a Sentiment Analysis via text mining and used YouTube Comment Scraper to retrieve a dataset from the YouTube platform. Then, the sentiment score and index of action were calculated from the review content and were regarded as dependent variables (i.e., emotion and purchase intention).
The results of the two-way ANOVA show that when the brand concept is functional, the non-humorous appeal induces positive emotion and higher purchase intention; whereas when the brand concept is symbolic, humorous appeal leads to positive emotion and higher purchase intention. Furthermore, emotion will mediate the interaction between humour appeal and brand concepts for predicting purchase intention.
With an understanding of how humour appeals affect emotion and purchase intention from text mining, marketers can better understand their target audiences and design more appropriate strategy in the online video more effectively. In addition, they could apply Sentiment Analysis to monitor messages form their audiences immediately.
|