Multivariate Network Visualization
A Multivariate Network (MVN) is an abstract data type that provides particular challenges for the information visualization research community. It permits the representation of complex relational data (stored in the form of a network) as well as the association of attributes with that data. The attributes themselves may use a range of different abstract data types. A MVN therefore consists of a set of objects, each of which has information associated with it. In addition, the objects are connected to each other in a network that represents the relationship between the objects. Further complexity is added when information is also associated with the inter-object relationships themselves. For example, a social network representation may consist of people (the objects), each of which has information associated with them (their age and post code).
MVNs prove particularly challenging for InfoVis researchers because of the wealth, richness and variety of the information that can be stored in them. The ISOVIS researchers have a strong interest on multivariate network visualization and focus on developing novel visual representations and interaction technologies for the visual analysis of MVNs. We also apply machine learning models and dimensionality reduction to create 2D embeddings that combine attributes and the network's topological information into one single unified model.