VISUALIZING PRODUCTIVE NETWORKS

Recommending new connections to users is an important tool for productive networks, but users are not always clear about how they are connected to the recommended contacts in the network graph. Providing
insight about the arguments used to build a recommendation increases the chances of its acceptance by users. This work is focused on recommendations that rely on an information model that enables recommendation systems, using only the relationships between users, items and keywords. These elements define the productive network graph, which this work represents in a visualization and interaction platform whose goal is twofold: to provide a discovery tool for productive networks, through which users can navigate the graph using different visualization contexts; and to provide insight about indirect relationships, representing which direct relationships are relevant for a particular recommendation. The platform is designed to work with any data that complies with this model. The visualization elements and interaction operations have been evaluated through the implementation of a case study, based on a sample of the Flickr network, producing positive results.