PDAV 2024

The increasing amount of data is a long-standing challenge for data analysis systems. Although building interactive systems has been a central focus of the visualization community, when applied to large-scale data and complex algorithms, most current visualization systems suffer from long, unmanaged computation delays between user interactions and system responses, rendering them unusable. The critical challenge we face here is to make a system’s latency manageable, ultimately ensuring it remains below the golden limits of human latency regardless of the amount of input data and complexity of algorithms.
Progressive Data Analysis and Visualization (PDAV) is a novel programming paradigm to control latency by replacing long computations with a series of smaller computations with bounded latency, improving iteratively until the whole computation is completed or until the user is satisfied with the latest iteration. Thus, PDAV computations also need to inform the user about the quality of the result to allow early decisions with controlled quality.
With PDAV, visual exploration systems can scale to large data sizes and use complex algorithms interactively, provided they are adapted to run progressively
The workshop will present state-of-the-art research and work-in-progress to design and implement PDAV systems.