DashGuide: Authoring Interactive Dashboard Tours for Guiding Dashboard Users
N. Hoque 1 and N. Sultanum 2 1 University of Iowa, Department of Computer Science, USA 2 Tableau Research, USA · 2025
Read DashGuide: Authoring Interactive Dashboard Tours for Guiding Dashboard Users by N. Hoque 1 and N. Sultanum 2 1 University of Iowa, Department of Computer Science, USA 2 Tableau Research, USA (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2504-17150.txt.
Introduction
Naimul Hoque and Nicole Sultanum present a 2025 system for making guided tours of data dashboards. Dashboards pack charts and controls that new users can find hard to read. Tours that walk users step by step help, but building them by hand eats time. Their system, called DashGuide, records what an author does on the dashboard and turns it into a playable tour.
Authors set a message goal, perform the steps, then trim each step with AI help. A pilot study with nine dashboard makers shaped the design. A fuller test with twelve makers found a good mix of speed, range, and creative room. Users rated the making process above older routes.
Catalog record
| CREATOR | N. Hoque 1 and N. Sultanum 2 1 University of Iowa, Department of Computer Science, USA 2 Tableau Research, USA |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 56 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2504-17150
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
DashGuide: Authoring Interactive Dashboard Tours for Guiding Dashboard Users N. Hoque 1 and N. Sultanum 2 1 University of Iowa, Department of Computer Science, USA 2 Tableau Research, USA Abstract Dashboard guidance helps dashboard users better navigate interactive features, understand the underlying data, and assess insights they can potentially extract from dashboards. However, authoring dashboard guidance is a time consuming task, and embedding guidance into dashboards for effective delivery is difficult to realize. In this work, we contribute DashGuide , a framework and system to support the creation of interactive dashboard guidance with minimal authoring input. Given a dashboard and a communication goal, DashGuide captures a sequence of author-performed interactions to generate guidance materials delivered as playable step-by-step overlays, a.k.a., dashboard tours. Authors can further edit and refine individual tour steps while receiving generative assistance. We also contribute findings from a formative assessment with 9 dashboard creators, which helped inform the design of DashGuide ; and findings from an evaluation of DashGuide with 12 dashboard creators, suggesting it provides an improved authoring experience that balances efficiency, expressiveness, and creative freedom. \ConferencePaper \SpecialIssuePaper \CGFccby Volume: 44 Issue: 3 ccs Human-centered computing Visualization systems and tools † † year: 2025 † † year: 2025 † † editors: W. Aigner, N. Andrienko, and B. Wang † † subject: Eurographics Conference on Visualization CGF Vol No † † teaser: Overview of DashGuide . a) The dashboard in this example visualizes different metrics from web traffic data (Sourced from Tableau Public [link] ). To create a guided tour for this dashboard, the dashboard author first selects “Dashboard Semantics and Encoding” as the communication goal for the tour (not shown) and then performs several consecutive click selections on the dashboard (shown with the hand cursors). DashGuide automatically generates content for the tour by prompting a large language model (LLM), which uses the communication goal and interactions provided by the author. The tour is then delivered as a step-by-step tour (b-f) where each step plays back the original interaction while placing the title and description for the step as an overlay near the location of interaction. The author can refine the content manually, provide further instructions to the LLM, or add or remove steps from the tour. 1 Introduction Visualization dashboards were conceived as tools for independent, self-service data exploration [ TBFC23 ] . In practice, however, dashboard users often require assistance in understanding components and functionality of a dashboard, from operating data filters to understanding the underlying data, interpreting charts, and navigating the dashboard to answer specific analytical questions [ CGM∗17 , CGM19 , KL16 ] . To this end, dashboard authors routinely provide guidance content to help these users, particularly those new to a dashboard [ DWH∗22 , CLL∗24 , SCW∗22 ] . Common forms of dashboard guidance include embedded text instructions and standard UI elements such as pop-ups and tooltips [ CGM∗17 , SCW∗22 , SS25 ] . While such guidance is relatively easy to author and often supported by authoring tools like Tableau, they tend to lack user engagement and may be skipped, missed or ignored by users [ SS25 , SWG∗21 ] . More recently, researchers have proposed interactive dashboard tours to improve user engagement and show diverse analytical capabilities of a dashboard [ DHF∗24 , LSS∗23 ] . These tours show usage scenarios, interactions, and insights in a step-by-step format, with each step incorporating a combination of content elements such as text descriptions and data highlights. Although recent work provides the tooling to instrument such tours [ DHF∗24 ] , authors still need to manually draft the content itself (e.g., text descriptions and images), which demands significant time and effort and offers limited scalability to supporting multiple tours with diverse usage scenarios. In view of these challenges, our work contributes: (a) a framework and system, DashGuide , to support dashboard creators in authoring dashboard tours and content more easily; (b) findings from a formative assessment that guided design and development of DashGuide ; and (c) findings from a qualitative user evaluation of DashGuide that underscore the value of our approach. Our user-centered design journey began with formative interviews with 9 dashboard practitioners. Interview findings point to time and effort being the biggest obstacles to crafting engaging and comprehensive guidance. While dashboard creators favor dynamic forms of guidance such as dashboard tours, technical limitations result in guidance being delivered in mostly static form (e.g., technical documentation external to the dashboard). Interviews also informed a range of communication goals for dashboard guidance, from demonstrating coordinated views to presenting key insights and explaining chart encodings. Based on this feedback, we designed DashGuide , our solution for easier authoring of interactive dashboard tours. To expedite the authoring process while accommodating author oversight and communication goals, DashGuide proposes an interaction-first approach. In this approach, a creator can quickly express their intent by choosing a high-level communication goal and then interacting with components of interest in the dashboard. DashGuide captures the interaction, compiles a sequence of key demonstration segments, and links interacted dashboard components and corresponding interactive actions with the respective segments. A Large Language Model (LLM) is then used to generate relevant content for each key segment, steered by the author’s informed communication goal. The dashboard tour is then rendered as a step-by-step guide where each step replays the corresponding interactions for the step, along with an explanatory title and description placed near the interacted components as overlays ( ). After an initial tour is created, DashGuide provides several edit and refinement features, including inserting new steps and regenerating content. We evaluated DashGuide in a user study with 12 dashboard creators where we asked participants to author dashboard tours for their own (or publicly available) dashboards. Qualitative findings suggest that DashGuide provided them with meaningful authoring support , and that the proposed interaction-first framework is a potentially promising approach for authoring dashboard tours. Our investigation also revealed open challenges around guidance authoring that future embodiments of our framework may help solve. 2 Related Work Our work builds upon prior works on tutorial systems in HCI, dashboard guidance, and data tours. We discuss these areas below. 2.1 Guidance and Tutorial Systems in HCI Guidance in HCI broadly encompasses problems and aids associated with learning to use software and hardware [ GF10 ] . The traditional approach to providing guidance is text documentation [ Car90 ] . However, documentation can be difficult to follow [ Ret91 ] , making them less accessible to users. A long line of HCI research has proposed solutions to make guidance more engaging and useful. The general consensus is that users prefer guidance that can show usage scenarios instead of outlining them in a static format [ GF10 , BCLO05 , CFPE21 , GAL∗09 , LGF12 ] . One popular approach is to adopt a step-by-step guided tour to show usage scenarios. Several works have proposed such tours for mobile applications [ ZLCL21 , CFPE21 ] , desktop and web applications [ CAR∗12 ] , and immersive environments (AR/VR) [ KNDH19 , HQW∗21 ] . Several open-sourced libraries (e.g., Intro.js [ Usa24 ] , TourGuide.js [ Tou24 ] , and WebTour.js [ Web24 ] ) are also available for creating guided tours. This