A study on Instagram user behavior of Instagram blogger post type by using sentiment analysis

• Main Authors: Chen, Yun-an, 陳韻安
• Other Authors: Lin, Hui-Fei
• Format: Others
• Language: zh-TW
• Published: 2018
• Online Access: http://ndltd.ncl.edu.tw/handle/7h35bg

碩士 === 國立交通大學 === 傳播研究所 === 106 === For the past few years, there are many researches focus on Instagram marketing. However the researches concerning bloggers and internet celebrities on Instagram are insufficient. Therefore, this study analyzed account of Instagram bloggers to explore if the different types of post will influence the Instagram user behavior.This study used sentiment analysis to analyze Instagram users’ comments. Mostly researches use questionnaires, content analysis in marketing research as the research methods .However it is difficult to avoid the subjectivity of the questionnaire design and limitation of sample size .And the content analysis is time-consuming. To solve few research limits as above, this study collected Instagram data by web crawling. Moreover, this study used these data to execute in sentiment analysis by using google natural language API. From this study, there are three results. First, the Instagram post types affect the behavior of Instagram users. Second, the number of likes and comments received from the video posts are more than the image posts. Lastly, In the aspect of sentiment scores of comments, video posts are also higher than the image posts. Therefore, this study proposes that the assumption that Instagram users have more positive attitude forward the video posts than the image posts.