Computational Linguistic and SNA to Classify and Prevent Systemic Risk in the Colombian Banking Industry

The banking sector has been one of the first to identify the importance of social media analysis to understand customers’ needs to offer new services, segment the market, build customer loyalty, or understand their requests. Users of Social Networking Sites (SNS) have interactions that can be analyzed to understand the relationships between people and organizations in terms of structural positions and sentiment analysis according to their expectations, opinions, evaluations, or judgments, what can be called collective subjectivity. To understand this dynamic, this study performs a social network analysis combined with computational linguistics through opinion mining to detect communities, understand structural relationships, and manage a Colombian case study’s reputation and systemic risk in the banking industry. Finagro and BancoAgrario are the network leaders in both centralities, most of the main actors have a negative polarity, and MinHacienda and cutcolombia with totally different orientations appear in all methods.