build interactive dashboards

Existing approaches struggle to reconstruct interactive data dashboards that support functionalities such as clicking and filtering. This work introduces Dashboard2Code, a novel task requiring models to actively explore interactive dashboards and integrate user interaction feedback to generate code that faithfully reproduces the target dashboard. To facilitate research in this direction, we present DashboardMimic, the first benchmark dataset built on Plotly+Dash, along with an automated evaluation framework that combines semantic analysis and dynamic interaction testing. Experimental results on 180 high-quality dashboard–code pairs demonstrate that current models exhibit limited performance on highly complex dashboards, with closed-source models significantly outperforming their open-source counterparts.

Existing visualization research predominantly focuses on *how to use* interactive features, neglecting the critical question of *how to construct* them. Method: We propose the first three-layer decoupled interaction authoring task model—intent–technique–component—derived from empirical coding and abstraction of 592 interaction units across 47 real-world applications. Contribution/Results: This model provides descriptive, evaluative, and generative capabilities, enabling the first unified formalization of interaction authoring intent, technical implementation, and component instantiation. It yields a reusable, theory-grounded classification framework that supports critical evaluation of existing visualization tools and informs the design and validation of next-generation low-code interaction authoring systems.

To address the challenges of low user engagement in dashboard onboarding and high authoring costs, this paper proposes a lightweight interaction-capture and generative refinement two-stage paradigm for communication-goal-driven automated guided tour generation. Methodologically: (1) natural interaction sequences are recorded and semantically parsed to automatically extract user actions and infer underlying intentions; (2) generative AI provides step-level editing assistance to refine tour content; and (3) overlay-based UI rendering enables playback of executable, stepwise tours. Our key contribution is the first integration of generative editing into dashboard tour authoring—balancing automation efficiency with expressive flexibility. Evaluation with 12 domain experts demonstrates significant improvements in tour creation efficiency and yields empirically grounded design principles for onboarding tools, now widely adopted in practice.

This work addresses the labor-intensive nature of presentation tasks—such as formatting and layout—in dashboard authoring, which currently lack support for partial reuse. Through a systematic user study, we characterize the needs and challenges associated with cross-source reuse of visual presentation elements. Building on these insights, we propose a novel paradigm that enables partial reuse of styles and layouts from multiple existing dashboards. We design and implement ReDash, a prototype system embodying this approach, and demonstrate through proof-of-concept experiments that our mechanism effectively overcomes key barriers in common reuse scenarios. The results show a significant improvement in authoring efficiency, confirming the feasibility and practical potential of partial reuse in real-world dashboard creation.

This work addresses the lack of effective evaluation mechanisms for large language models (LLMs) in generating interactive analytical dashboards, which hinders the assessment of their analytical reasoning and interaction quality. The authors propose the first open-ended benchmark for interactive dashboard generation, requiring models to produce both a functional dashboard and a replayable interaction trace. Evaluation is automated through a browser-based executor and a vision-language model (VLM), with the interaction trace serving as a novel basis for judgment. A lightweight, open-source judge model, DashJudge-8B, combined with Bradley-Terry aggregation, enables scalable leaderboard construction. Experiments demonstrate that DashJudge-8B reliably replicates human judgments, that interaction evidence substantially improves evaluation consistency, and that state-of-the-art models still exhibit significant shortcomings in rendering fidelity, analytical depth, and interactive behavior.

Contemporary dashboards suffer from complex interactions and tightly coupled views, necessitating labor-intensive authoring and maintenance of guided tutorials—resulting in high costs and poor synchronization with dashboard updates. To address this, we propose DIANA, the first multimodal dashboard assistant integrating speech, text, and mouse gaze inputs. Built upon large language models (LLMs), DIANA implements a context-aware, real-time interactive system that enables users to issue queries via any combination of modalities and receive immediate visual feedback—including interface element highlighting—and semantically grounded explanations. Its key innovation lies in transcending the conventional unimodal text-based LLM paradigm by enabling synergistic, tri-modal-driven dynamic guidance. DIANA establishes, for the first time in visualization analytics, a closed-loop pipeline spanning multimodal input, interface response, and semantic output. A user study demonstrates that DIANA significantly improves users’ comprehension efficiency of dashboard structure and functionality while markedly reducing reliance on manual guidance.

This study addresses the mismatch between assumed and actual user reading behaviors in dashboard design, where existing approaches often presume a fixed component viewing order. Through a mixed-methods investigation involving 18 designers and 16 users, the work systematically identifies and quantifies six key factors influencing interaction sequences: layout, visual salience, semantics, functional role, interactivity, and user context. Integrating interviews, behavioral logs, and sequential analysis, the research uncovers both consistent patterns and diverse strategies in how users navigate dashboards, revealing representative navigation pathways. These findings provide an empirical foundation and actionable design insights for computational modeling of dashboard comprehension and the development of intelligent guidance systems that adapt to real user behavior.

Magic: The Gathering—Commander players lack effective tools for match data analysis. Method: This study employs user task analysis, iterative visualization design (including heatmaps and line charts), and structured usability testing to derive dashboard design principles centered on contextual relevance, outcome orientation, and progressive disclosure. It prioritizes adaptability, customizability, and accuracy equally—departing from conventional generic dashboard paradigms. Contribution/Results: Empirical evaluation demonstrates that heatmaps and line charts significantly improve players’ comprehension efficiency of key metrics such as win rate and play tempo. Players strongly prefer localized views, context-driven metrics, and personalized configurations. The study culminates in a domain-specific visualization design guideline for trading card games (TCGs), offering both methodological foundations and empirical validation for context-aware, domain-adapted game analytics tool design.

This work addresses the state desynchronization between natural language queries and dashboard interactions in multi-step business intelligence (BI) analysis by proposing the first agent-based digital twin framework. The approach couples large language model (LLM) agents with executable dashboard states and reconstructs a shared analytical context through unified interaction logs, thereby ensuring consistency across dialogue, user actions, semantic alignment, and provenance tracking. Innovatively leveraging a digital twin mechanism, the framework enables state-aware analytical summarization and traceable context management. Experimental results demonstrate a 20.0% absolute improvement in exact-match accuracy (from 43.3% to 63.3%), a partial-match accuracy of 70.8%, and a reduced timeout rate of 10.0%. User studies further confirm high task accuracy and a positive interactive experience.

This work addresses the limitations of traditional real-time analytics systems, which rely on manually defined queries and struggle to proactively uncover the vast array of potential insights within complex, dynamic data streams. To overcome this, the authors propose a multi-agent architecture that establishes a continuous closed-loop process for autonomous insight discovery, encompassing hypothesis generation, compilation of executable analyses, result validation, and visualization. A key innovation is the introduction of a contract-driven design based on typed intermediate artifacts, which ensures modularity, observability, lineage tracking, and secure execution of dynamic analyses. The system leverages Kafka as its event coordination backbone and Flink for stream processing, integrating large language models to power specialized agents. Empirical evaluations in retail, financial, and public data scenarios demonstrate an effective paradigm shift from query-driven to proactive discovery-driven analytics.