A declarative domain independent approach for querying and generating visualizations
Constructing visualizations using modular visualization environments (MVEs) requires knowledge about visualization transformation theory as well as implementation specific details about how to programmatically chain together relevant sets of modules into pipe and filter-like architectures known as pipelines. Constructing visualization pipelines introduces a number of challenges including: understanding how to transform raw data into forms that can be ingested by MVEs, identifying target modules that map data into graphical data, and understanding how to further transform resulting graphical data into forms that can be presented. Users typically must immerse themselves in a deluge of documentation and usage examples before becoming proficient in the aforementioned tasks and therefore able to generate suitable visualizations required for analysis. This dissertation presents a visualization query language that was designed with the goal of hiding the associated complexities of visualization pipeline construction by abstracting away implementation specific details. By posing visualization queries, users can declaratively specify what visualizations they want generated from their data and quickly explore a variety of results that may only be generated using complex combinations of modules possibly supported across different MVEs. This dissertation explores the query answering facilities associated with visualization queries as well as presenting a pilot study which supports the claim that users both prefer and are more proficient at using queries than writing traditional pipeline-based code.
Information Technology|Computer science
Del Rio, Nicholas, "A declarative domain independent approach for querying and generating visualizations" (2013). ETD Collection for University of Texas, El Paso. AAI3594331.