Provenance for Interactive Visualizations
Provenance for Interactive Visualizations
Fotis Psallidas, Eugene Wu, Proceedings of the Workshop on Human-In-the-Loop Data Analytics. ACM, 2018.
This paper explore the connection between the interactive visualizations and data provenance.
Interaction as provenance
It starts from core interaction logic with provenance: selections, logic over selections and multi-view linking. Interactive selections is to get subset of inputs that corespond to selected visual outputs, for example, finde the airports that operate at the selected states. Logic over selections is to express application logic over the selected inputs, for example, find the #airports that operate at the selected states. Multiple-view linking is to look at the relationships between different views, for example, show the distribution of #flights per carrier only for selected states shown as below.
Interactive selections such as item selection, group selection, range selection and generalized selections can use a common provenance operation known as
backward_trace()
Logic over selection such as tooltips, details-on-demand and semantic zooming can be expressed as queries that take the backward_trace() of the user’s selection as input:
SQL(backward_trace(…))
Multiple-link such as linked brushing and crossfilter can be expressed by the backward_trace() followed by refreshing the other views by executing the queries shown as below. Selective refresh is to update procedure corresponds to a common provenance operation.
selective_refresh(backward_trace(…))
Multi-application linking
Many applications are built over the same database or dataset. Extend multi-view linking to multi-application linking is possible, and it can connect data across applications and reuse the applications logic. The figure below shows linking between the running visualization with external applications.
