interactive data inspection

Existing visualization tools commonly embed data loading and transformation logic within visualization components, leading to redundant development efforts, high user learning costs, and difficulties in cross-tool data interoperability. Method: This paper proposes a modular architecture that systematically decouples data processing from visualization logic for the first time. It defines standardized data interfaces and a dynamic integration mechanism, implemented as a web-based prototype supporting bidirectional data flow and parallel collaboration across heterogeneous tools while preserving their autonomy. Contribution/Results: The architecture establishes a unified data interaction layer without compromising tool independence. Empirical evaluation demonstrates significant reductions in developers’ reimplementing overhead and users’ learning curves. Moreover, it enables scalable, open, and collaborative visualization ecosystems—introducing a novel, extensible paradigm for integrated visual analytics.

This work addresses the limitations of linear conversation logs generated by existing conversational data analysis systems, which hinder data workers’ ability to retrospect and communicate about nonlinear, iterative analytical processes. To overcome this, the paper proposes a structured dialogue presentation method that introduces probes enabling multi-level navigation, on-demand detail expansion, and context-enhanced summarization—going beyond conventional scrolling and keyword search. By integrating visual recall with sequential and abstraction-based navigation strategies, the approach effectively supports users in recalling, reorienting within, and prioritizing past analytical exchanges. A user study with ten participants demonstrates that the method significantly enhances traceability of analytical reasoning and improves collaborative efficiency, validating its effectiveness in real-world data analysis workflows.

Life sciences face significant challenges due to the limitations of conventional static databases in supporting exploratory querying, real-time analytics, and multidimensional dynamic visualization. To address these issues, this paper proposes a user-centric interactive database framework that integrates modern data management architectures, scalable storage engines, reactive front-end visualization, and ontology-driven data standardization. For the first time, the framework systematically incorporates authentic research workflows—such as cell-line screening—thereby unifying data generation, biological interpretation, experimental design, and clinical correlation. The system enables high-concurrency, low-latency real-time queries and cross-modal (e.g., genomic, imaging, clinical) integrated dynamic analysis. Empirical evaluation demonstrates substantial improvements in exploratory data analysis efficiency and reproducibility of scientific findings.

To address the dual challenges of insufficient personalization and low efficiency in interactive exploration for automated insight discovery, this paper proposes InsightMap—a map-metaphor-based framework for insight visualization and hybrid discovery. Methodologically, it formalizes data insights as measurable, layout-aware data objects; introduces a similarity metric integrating semantic and statistical features; and establishes a hybrid paradigm that synergistically combines automated mining with interactive exploration. InsightMap enables seamless transitions from global overviews to localized deep-dive analysis. Through multiple case studies and user experiments, InsightMap reduces average task completion time by 37% and achieves a user satisfaction score of 4.8/5.0, demonstrating significant improvements in both insight discovery efficiency and personalized adaptability.

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.

This study addresses the unclear mechanisms underlying the creative processes and expressive form evolution of data artists. It presents the first systematic deconstruction of the end-to-end data art creation workflow through a multi-phase investigation integrating project case analyses, in-depth interviews, and task-based design experiments. The research reveals the co-evolutionary dynamics between narrative and visual forms, elucidates the pivotal roles of inspiration sources and sketch prototyping, and identifies key design trade-off strategies employed by practitioners. These findings provide a theoretical foundation and actionable design guidelines for developing AI-assisted visualization tools that support creative decision-making in data art practice.

This study addresses the limitation of existing data analysis agents that overlook exploratory phases, resulting in inadequate comprehension of complex tabular data. We propose establishing data exploration as a primary evaluation objective for LLM-based analysis. By constructing a multi-table workbook benchmark and extending DSBench, we validate this approach through structured artifact evaluation. Experimental results demonstrate that explicit data exploration compensates for logical structural deficiencies in strong models, significantly improving downstream task accuracy and human-AI collaboration efficiency. These findings confirm the critical value of data exploration as an independent, inspectable stage and a verification node in human-machine workflows, thereby offering a novel paradigm for reliable data analysis.

This study investigates how interface paradigms of data cleaning tools shape users’ actual cleaning strategies. Through a between-subjects observational experiment with 40 participants, it compares usage behaviors across Jupyter, Excel, ChatGPT, and OpenRefine on representative data cleaning tasks, applying for the first time the technical dimensions framework from programming systems to this domain. Findings reveal that interface design significantly steers user strategies without determining outcomes: data-centric interfaces (e.g., Excel) encourage opportunistic, ad-hoc operations, whereas abstraction-centric tools (e.g., Jupyter) facilitate systematic transformations at the cost of higher cognitive load. The results uncover systematic trade-offs among tools, indicating no single optimal choice and underscoring the critical role of interface design in shaping data work practices.

Existing large language model (LLM)-driven data analysis tools are often confined to isolated subtasks and struggle to support end-to-end executable analytical workflows. This work proposes an autonomous, sandboxed, and auditable end-to-end system that leverages LLMs for action planning, iteratively generating structured operations, executing code in a secure environment, and integrating streaming traceability with intermediate result previews. By unifying a structured action backend, sandboxed execution, and an interactive visual interface—features integrated here for the first time—the system enables users to drive complete analytical workflows using only natural language. Users can inspect, modify, and export the entire process and its outputs directly within a web browser, ensuring full reproducibility, editability, and transparency throughout the analytical pipeline.

This work addresses the disconnect between automated and manual verification in existing fact-checking systems, which hinders readers’ ability to flexibly collaborate with verification tools during reading. The authors propose FYI, a browser extension that embeds fact-checking directly into the reading environment and offers four complementary tools spanning fully automated to manually driven exploration. FYI establishes a collaborative verification paradigm centered on visualization, positioning AI as a starting point rather than an authoritative source. An exploratory study with 22 participants revealed three prevalent user workflows—AI-first, manual-first, and parallel co-auditing—and demonstrated that visualization serves as a critical mechanism for auditing AI-generated claims. Trust in the system increased when multiple tools converged on consistent results and decreased when they diverged. The system is open-sourced to advance research in hybrid verification.