← All publications

Dissertation Guidance & Analytic Provenance

Designing, Developing, and Democratizing Guidance for Visual Analytics

Arpit Narechania

Georgia Tech · 2024

★ 2025 IEEE VGTC Visualization Dissertation Award

Teaser for Designing, Developing, and Democratizing Guidance for Visual Analytics

Abstract

The ubiquity and utility of data across a variety of domains have created an urgent need for automated systems that help users make sense of complex information. While these systems can process vast amounts of data, human intuition and expertise remain critical for many tasks, necessitating effective collaboration between the two for accurate and timely decision-making. However, challenges arise when human users of these systems must provide extensive input (e.g., to convey their analytic intent) or when automated actions by systems misinterpret user intent or are mistimed, which can increase users’ perceptual and cognitive load and disrupt the analytic process. Guidance (or any kind of help) offers a promising solution to bridge this knowledge gap between human expertise and (humans’ understanding of) system capabilities, while improving the quality and effectiveness of the analysis process and its outcomes. This dissertation deepens our understanding of how guidance can be communicated to/from users and how it can impact users’ behavior during analysis, with broader implications for researchers, developers, and practitioners via three thrusts: (1) Design: design spaces for provenance and guidance communication, derived from a series of design interventions for guiding users during various analysis tasks; (2) Develop: guidance-enriched systems, developed for and evaluated with end-users, revealing strengths and challenges, and informing future systems; and (3) Democratize: an open-source library of guidance-enriched user interface controls, helping developers prototype custom systems.

Citation

@phdthesis{narechania2024guidance,
  author = {Narechania, Arpit Ajay},
  title = {{Designing, Developing, and Democratizing Guidance for Visual Analytics}},
  school = {Georgia Institute of Technology},
  year = {2024},
  month = dec,
  type = {Ph.D. dissertation},
  doi = {https://hdl.handle.net/1853/76976},
  url = {https://hdl.handle.net/1853/76976}
}