Description
Social media managers as well as analysts use social media messages as an additional channel to manage and measure opinions towards brands. Currently, however, sentiment tools have been predominantly focused on the binary positive-negative valence of emotions and thereby neglected the multi-dimensional structure of human emotions. Moreover, while practitioners try to address the issue of undifferentiated customer requests towards a brand by operating more interest group-specific accounts, research still lacks understanding regarding the impact of different account types. We approach this research gap by developing a classification of social media accounts and, subsequently, deploy a sentiment analysis that differentiates between seven emotions within 532,363 thousand tweets towards 641 accounts from 33 S&P 100 companies. Our results confirm the assumed necessity of considering different account types when studying corporate social media presence and assessing differentiated emotions in social media analytics.
Recommended Citation
Risius, Marten and Akolk, Fabian, "Differentiated Sentiment Analysis of Corporate Social Media Accounts" (2015). AMCIS 2015 Proceedings. 7.
https://aisel.aisnet.org/amcis2015/SocialComputing/GeneralPresentations/7
Differentiated Sentiment Analysis of Corporate Social Media Accounts
Social media managers as well as analysts use social media messages as an additional channel to manage and measure opinions towards brands. Currently, however, sentiment tools have been predominantly focused on the binary positive-negative valence of emotions and thereby neglected the multi-dimensional structure of human emotions. Moreover, while practitioners try to address the issue of undifferentiated customer requests towards a brand by operating more interest group-specific accounts, research still lacks understanding regarding the impact of different account types. We approach this research gap by developing a classification of social media accounts and, subsequently, deploy a sentiment analysis that differentiates between seven emotions within 532,363 thousand tweets towards 641 accounts from 33 S&P 100 companies. Our results confirm the assumed necessity of considering different account types when studying corporate social media presence and assessing differentiated emotions in social media analytics.