Abstract

The individuals in a network tend to form homophilic clusters with their peers, with some of them acting as strong connectors (hubs) between distinct groups or single nodes. Specific network centrality metrics quantify the role of any node in a network by incorporating the number of connections, the density of a region, and the weight of connections. Recent studies have investigated network’s multi-dimensionality as the source of information for each node. However, this pertains to one-mode networks that include one type of actor, and little is known about protagonistic roles in two- or multi-mode networks with two or more types of actors, respectively. Here we propose a novel methodology that uses a vector-based multi-space model and quantifies the complete information for the elements in a network, considering their multiple interactions. We applied our methodology to a document-based network of travel guides and a literature-based network of artists, identified the protagonists, and evaluated their protagonistic role. Our innovative methodology could measure the multi-dimensional relationships in various real-world systems and reveal the key actors and interactions.