Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
119 tokens/sec
GPT-4o
56 tokens/sec
Gemini 2.5 Pro Pro
43 tokens/sec
o3 Pro
6 tokens/sec
GPT-4.1 Pro
47 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Data Navigator: An accessibility-centered data navigation toolkit (2308.08475v1)

Published 16 Aug 2023 in cs.HC

Abstract: Making data visualizations accessible for people with disabilities remains a significant challenge in current practitioner efforts. Existing visualizations often lack an underlying navigable structure, fail to engage necessary input modalities, and rely heavily on visual-only rendering practices. These limitations exclude people with disabilities, especially users of assistive technologies. To address these challenges, we present Data Navigator: a system built on a dynamic graph structure, enabling developers to construct navigable lists, trees, graphs, and flows as well as spatial, diagrammatic, and geographic relations. Data Navigator supports a wide range of input modalities: screen reader, keyboard, speech, gesture detection, and even fabricated assistive devices. We present 3 case examples with Data Navigator, demonstrating we can provide accessible navigation structures on top of raster images, integrate with existing toolkits at scale, and rapidly develop novel prototypes. Data Navigator is a step towards making accessible data visualizations easier to design and implement.

User Edit Pencil Streamline Icon: https://streamlinehq.com
Authors (3)
  1. Frank Elavsky (6 papers)
  2. Lucas Nadolskis (2 papers)
  3. Dominik Moritz (36 papers)
Citations (15)

Summary

Overview of "Data Navigator: An Accessibility-Centered Data Navigation Toolkit"

The paper "Data Navigator: An Accessibility-Centered Data Navigation Toolkit," authored by Frank Elavsky, Lucas Nadolskis, and Dominik Moritz, addresses prevalent challenges in making data visualizations accessible for people with disabilities. The research focuses on the deficit in current visualizations, which often remain inaccessible due to reliance on visual-only rendering, lack of navigable structures, and inadequate support for diverse input modalities.

Data Navigator is presented as a novel toolkit aimed at alleviating these accessibility barriers. Built upon a dynamic graph structure, it facilitates the creation of navigable interfaces encompassing lists, trees, graphs, and spatial or geographic relations. This system distinguishes itself by supporting a wide array of input modalities beyond traditional mouse and touchscreen, including screen readers, voice commands, keyboards, gesture detection, and DIY fabricated devices.

Technical Pivot: Graph Structures

At the core of Data Navigator is its graph-based approach to data visualization. Graphs, due to their generic nature, provide a flexible substrate capable of representing various data structures such as lists, trees, and more complex relational or spatial forms. This flexibility supports the construction of intuitive navigation structures that map more closely to the semantic relationships within the data. By decoupling visual elements from navigation structures, Data Navigator allows developers to build interfaces where accessibility elements can be dynamically controlled, ultimately enhancing the navigability of visualizations.

Input Modalities and Interaction

Data Navigator emphasizes abstract navigation, enabling input modality independence. This design philosophy allows multiple methods for interaction, such as voice input or gestures, to trigger navigation events within the software. Additionally, non-pointer-based devices, commonly used by individuals with dexterity impairments, are considered, expanding the toolkit's applicability.

By incorporating sequential, discrete navigation tactics, akin to keyboard navigation, Data Navigator opens new possibilities for designing accessible data exploration experiences. This approach ensures that devices leveraging assistive technologies can seamlessly interact with visualization content.

Rendering and Semantic Flexibility

The toolkit allows loose coupling between navigable structures and rendered visuals, offering developers the freedom to customize the rendering process. This flexibility ensures that focus indicators can be added even in raster-based images, which are traditionally non-interactive, thus expanding the toolkit's potential applications. Developers can opt for on-demand rendering of accessibility elements, maintaining efficiency and reducing memory overhead.

Case Studies and Implications

The paper demonstrates Data Navigator's practical implementation through diverse case studies, including augmenting static visualizations and integrating with existing visualization ecosystems like Vega-Lite. These examples highlight the toolkit's ability to extend accessibility to raster images and canvas-rendered graphics, which generally lack inherent accessibility features.

The implications of this work extend to both theoretical and practical domains. Theoretically, Data Navigator contributes a new perspective on structuring navigable data interfaces, aligning with broader accessibility guidelines. Practically, it provides a substantive toolset for researchers and developers aiming to remediate existing visualization tools and create new, accessible designs.

Future Developments

Potential future directions include enhancing the automatic generation of graph structures from visualization datasets, refining the scalability of navigation across extensive datasets, and expanding support for emerging input technologies. Engaging with a broader user base, especially with diverse disabilities, could further refine the toolkit's utility and inform subsequent iterations.

In summary, "Data Navigator: An Accessibility-Centered Data Navigation Toolkit" contributes a significant advancement in accessible data visualization interfaces, offering both a theoretical framework and a practical toolset poised to impact the design and development of more inclusive visual data experiences.

Youtube Logo Streamline Icon: https://streamlinehq.com