The reactive vega stack : declarative interaction design for data visualization
Abstract/Contents
- Abstract
- Interactive visualization is an increasingly popular medium for analysis and communication as it allows readers to engage data in dialog. Hypotheses can be rapidly generated and evaluated in situ, facilitating an accretive construction of knowledge and serendipitous discovery. Crafting effective visualizations, however, remains difficult. Programming toolkits are typically required for custom visualization design, and impose a significant technical burden on users. Moreover, existing models of visualization relegate interaction to a second-class citizen: imperative event handling callbacks that are difficult to specify, and even harder to reason about. This thesis introduces the Reactive Vega stack: two new declarative languages for interactive visualization that decouple specification (the what) from execution (the how). At the foundation is Reactive Vega, an expressive representation that models user input as streaming data. Its underlying dataflow runtime handles the complexity of event propagation and state management, freeing users to focus on interaction design decisions. Vega-Lite builds on Vega to provide a high-level grammar for rapidly authoring interactive graphics for exploratory analysis. Its concise specification format decomposes interaction design into semantic units that can be systematically enumerated. Critically, these languages offer JSON syntaxes to simplify programmatic generation of interactive visualization and enable novel interactive data systems. This thesis develops one such system, Lyra, a direct manipulation tool for visualization design. Drag-and-drop operations in Lyra generate statements in Vega and Vega-Lite, allowing users to author a diverse range of visualizations without any textual specification. These systems have been released as open-source projects, have been widely adopted, and have given rise to an ecosystem of interactive visualization tools. Users can author an exploratory visualization in the Jupyter Notebook, export it to Lyra via Vega-Lite and add an explanatory annotation layer, and then embed the resultant Reactive Vega visualization within a Wikipedia article. As a result, rather than a single monolithic system, the Reactive Vega stack facilitates development of targeted applications, and allows users to work at the level of abstraction most suited for the task at hand.
Description
| Type of resource | text |
|---|---|
| Form | electronic; electronic resource; remote |
| Extent | 1 online resource. |
| Publication date | 2017 |
| Issuance | monographic |
| Language | English |
Creators/Contributors
| Associated with | Satyanarayan, Arvind |
|---|---|
| Associated with | Stanford University, Computer Science Department. |
| Primary advisor | Agrawala, Maneesh |
| Primary advisor | Heer, Jeffrey Michael |
| Thesis advisor | Agrawala, Maneesh |
| Thesis advisor | Heer, Jeffrey Michael |
| Thesis advisor | Landay, James A, 1967- |
| Advisor | Landay, James A, 1967- |
Subjects
| Genre | Theses |
|---|
Bibliographic information
| Statement of responsibility | Arvind Satyanarayan. |
|---|---|
| Note | Submitted to the Department of Computer Science. |
| Thesis | Thesis (Ph.D.)--Stanford University, 2017. |
| Location | electronic resource |
Access conditions
- Copyright
- © 2017 by Arvind Satyanarayan
- License
- This work is licensed under a Creative Commons Attribution 3.0 Unported license (CC BY).
Also listed in
Loading usage metrics...