Towards an Analytical Probe for Twitter Information Flow Micro-structure

This article reviews the concepts presented with the conceptual framework of the network society, both in historical and present considerations, and extends them towards creating an analytical probe for information flow micro-structure on Twitter. By considering the effects of societal interconnectedness as it relates to individuals, the philosophical foundations of the network society are predicated on network interconnectedness that goes beyond institutional integration. Although the manifestations are progressively becoming apparent, the literature is still largely lacking in empirical results to facilitate further theoretical developments. Specifically; we present arguments that support the notion that society is becoming highly interconnected, and that these developments warrant study. As a technical contribution, we present the foundations for an analytical tool that is formulated from the joint application of machine learning and network science. Technically, we construct a multilayer network composed of: user interaction patterns, topics that characterize the discourse, and the entities that are referenced in the discourse. The multilayer network then forms the mathematical object on which analysis can be conducted, and allows us to study information flow micro-structure. The set of multilayer networks, that is generated by varying time, serves as a technical underpinning on which to characterize information flow micro-structure on Twitter. As an overarching theme, we argue that developing an understanding of such information flows is of significance to: commercial entities, governments, and the general populace.