Information Networks to Derive Value from Social Media

The rise in electronic interactions has made information networks ubiquitous. Correspondingly, research across multiple domains has begun to acknowledge the social and economic value of these networks for business decision-making. In this dissertation, we derive brand networks from social media to provide statistical knowledge on online market structures and automatically infer brand associations over time. Compared to extant data mining approaches that rely on substantial human intervention, this unsupervised automated approach lets practitioners study the relative positioning of their brand not only against a set of common competitors but against any other brand in the ecosystem; thus, uncovering a broader picture on both within-industry competition and across-industry complementarities. To investigate the usefulness of our proposed methodology, we validate the findings from our automated approach against external survey ratings and conduct extensive robustness checks to ensure reliability of underlying Twitter data. Large scale data focused methods for brand management are relatively new and present many opportunities for future research. We hope the methods introduced in this dissertation serve as a foundation for researchers interested in leveraging implicit brand networks for gaining insights into consumers and brands.