Text mining of social media data: insights and implications related to an enterprise data breach incident

Purpose The rise of the internet has made data breaches one of today’s top security concerns. Although there is an increasing number of studies on data breaches, few have adopted the perspective of individual victims to help develop effective strategies and mitigate the undesirable consequences. This study aims to uncover the emotional and thematic content of discussions on social media after an enterprise data breach incident and to explore how these elements influence information diffusion. Design/methodology/approach The authors extract 16,868 posts related to a data breach incident involving China’s Didi ride-hailing firm and conduct text mining (emotion detection and topic modeling). Negative binomial regression is used to examine how emotions and topics influence information diffusion. Findings The results reveal two predominant emotions – disgust and joy – differing from prior findings that primarily identified anger and other negative emotions. The six topics distilled from the posts are about punishment, approval, risks, blame, data security and platform. Further analyses reveal that negative emotions and topics on risks and data security are positively related to information diffusion. Practical implications The results of this study provide support for regulators to have stricter enforcement of laws and regulations on privacy security. Practitioners should also pay attention to emotions and topics on social media, as these factors can influence information diffusion. Originality/value This study is among the few to provide insights into individual victims’ reactions to a data breach incident in a non-Western context.