Скачать или смотреть Interactive Visualization of Provenance Graphs for Reproducible Biomedical Research

Poster @ IEEE InfoVis'15
Authors: Stefan Luger, Holger Stitz, Samuel Gratzl, Nils Gehlenborg, Marc Streit
Abstract: A major challenge of data-driven biomedical research lies in the collection and representation of provenance information to ensure reproducibility of the gained results. The Refinery Platform is an integrated data management, analysis, and visualization system designed to support reproducible biomedical research. In order to communicate and reproduce multi-step analyses on datasets that contain data of hundreds of samples, it is crucial to be able to visualize the provenance graph at different levels of detail. Most existing approaches are based on node-link diagrams, however, they usually do not scale well to large graphs. Our proposed visualization technique dynamically reduces the complexity of subgraphs through hierarchical aggregation and application of a degree-of-interest (DOI) function to each node. Triggered by user interactions, such as selecting a subset of analyses or a path in the graph, unaffected parts of the graph are dynamically aggregated into a glyph representation. We further reduce the complexity of the provenance graph visualization by layering identical or similar sequences of parallel analysis steps into an aggregated sequence.