Bind and Adapt - A paradigm and framework for information visualization

Denis Lalanne
Benoît Pointet
Project status: 

This master thesis proposes to reconsider the visualization process by focusing of the relationships between data and its visual representations. A novel `Bind & Adapt' approach to visualization is presented, which considers data and visuals as partners in a structured process of bindings and adaptations, operating the visualization and supporting interaction upon it.

Reflecting this approach, a declarative language and ad hoc interpreter have been developed and integrated into the Metaphorea framework, a prototypal web-native declarative framework for visualization, which bases on the principles of declarative expression of visual mapping, reuse of visual forms, clear distinction and parity of the data and visual realms.

Through the lens of four use cases, this thesis discusses the approach and the framework, highlighting strengths and weaknesses. Key achievements are a unified and flexible model for both data and its representations, and a unified declarative language to express key features of a visualization such as data manipulation (aggregation, filtering, partitioning), visual mapping, and interaction coordination.

Further developments of the Metaphorea framework are proposed such as direct manipulation interface, view composition and generation of guides and scales, so to complete the fulfillment of the Bind&Adapt approach.