Pith. sign in

REVIEW 4 major objections 5 minor 73 references

VA-Blueprint: Uncovering Building Blocks for Visual Analytics System Design

T0 review · 4 major / 5 minor · reviewed 2026-08-05 · deepseek-v4-flash

Pith's one-line read Visual analytics systems decompose into a three-level blueprint of reusable building blocks, and LLM extraction from 101 papers can build that catalog at scale.

desk verdict A useful, honest knowledge base and extraction pipeline, but the text-only LLM scaling puts the temporal trends on shaky ground. read the letter →

arxiv 2508.07497 v1 pith:XYUICOO7 submitted 2025-08-10 cs.HC cs.AI

classification cs.HCcs.AI
keywords visualanalyticsknowledgebasesystemdesignbuildingblocksdataflowmodellargelanguagemodelsurbantaxonomy
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

VA-Blueprint is a methodology and knowledge base for visual analytics (VA) system design. The paper claims that any VA system can be represented as a hierarchical dataflow of building blocks — high-level stages, intermediate functional groups, and granular operations — connected by data and interaction dependencies. The authors manually coded 20 urban VA papers to fix this schema, then used a large language model with human review to scale extraction to 101 papers, producing a machine-readable catalog of more than 2,500 components and 2,400 dependencies. Interviews with system authors and a quantitative comparison against manual annotation support the claim that the extracted blueprints faithfully capture the architecture, while also exposing a known weakness: papers omit some functionality, so the knowledge base inherits those gaps. If the claim holds, developers gain a queryable catalog of proven component configurations instead of assembling bespoke pipelines from scratch.

What carries the argument

The load-bearing mechanism is the hierarchical JSON blueprint schema (SystemBlueprint): each system is a directed graph whose nodes are GranularBlocks nested inside IntermediateBlocks nested inside HighLevelBlocks, each granular block carrying a unique ID, a quoted source citation, inputs, outputs, and a FeedsInto edge list; edges are typed as data dependencies or interaction dependencies. This single schema simultaneously structures LLM output, enables machine-readable comparison, and drives a visual node-link interface for validation.

What would settle it

Pick a set of VA systems with publicly available source code and reconstruct each system's component graph directly from that code; compare against the paper's blueprint for the same system. If the code-grounded reconstruction consistently finds meaningful components or dependencies absent from the paper-based blueprint (beyond the 7 blocks and 5 edges the expert interviews already found), the completeness premise fails. A cheaper check: rerun the LLM extraction on the same 20 papers the authors manually coded and test whether the missing Data Processing blocks are present in the papers' figur

Watch

Extended reading notes

Core claim

VA systems have a describable, shared anatomy: a directed graph of components organized into three abstraction levels (high-level stages, intermediate groups, granular blocks) with edges carrying data or interaction dependencies. The authors show this structure is stable enough to serve as a strict JSON schema that guides both manual coding and LLM extraction; across 101 urban VA papers it yields on average 25 components and 24 dependencies per system, with Visualization the most common and central component class. The paper also reports that an LLM, prompted with the schema plus few-shot examples and corrected by human review, matches manual labels at an 86% rate and cuts annotation time fr

Load-bearing premise

The knowledge base is built entirely from research papers — text and figures — rather than from source code or running systems; if papers routinely omit or ambiguously describe their actual functionality, the blueprints and the trends derived from them inherit those gaps.

Editorial extensions

If this is right

  • Developers can query the knowledge base for common component configurations (e.g., Map 2D plus Area Selection) and reuse proven architectures instead of designing pipelines from scratch.
  • VA systems can be compared quantitatively: complexity, hub components, and interaction loops become measurable graph properties.
  • The temporal trends the paper reports (component count rising from 13.5 in 2007 to 15.8 in 2024; edge count from 21.8 to 30.4) make the knowledge base an empirical record of how VA architectures evolve.
  • The taxonomy can ground future low-code authoring tools and model-driven development, with blueprints serving as formal specifications.
  • LLM-assisted extraction with human review cuts annotation effort from 18+ hours to under 2 hours per batch, making it practical to update the knowledge base as new systems appear.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The same methodology should transfer to other VA domains (biology, health, climate), but the component vocabulary and trend numbers are urban-specific until replicated elsewhere.
  • Because papers omit functionality (experts found 7 missing granular blocks and 5 missing edges across 9 systems), the literature-only ceiling suggests that fusing source code, manuals, or live demos into the pipeline is the natural next step, and the JSON schema is well suited to such fusion.
  • The LLM's tendency to under-extract Data Processing blocks by 22% suggests future pipelines should either spend more review effort on processing components or add diagram/figure understanding to the extraction.
  • The blueprint graphs could be mined for recurring design archetypes (e.g., filter-centric hubs vs. progressive-decomposition loops), providing a design-pattern catalog the paper only begins to sketch.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper introduces VA-Blueprint, a methodology and knowledge base for documenting the building blocks of urban visual analytics (VA) systems. The authors curate a corpus of 101 papers, manually analyze 20 papers to develop a three-level JSON schema (high-level, intermediate, and granular blocks, with data and interaction dependencies), and then use GPT-4 with few-shot prompting and human-in-the-loop review to extract blueprints for the remaining 81 papers. The resulting knowledge base is publicly available and includes 403 high-level, 756 intermediate, and 1,434 granular blocks with 2,448 edges. The paper analyzes component frequencies, network topology, temporal evolution, and three example systems, and evaluates the resource through five expert interviews and a quantitative comparison between LLM and manual annotations on 20 held-out papers. The main claims are that the method systematically organizes VA system components and that LLM-assisted extraction provides a strong baseline that still requires expert refinement.

Significance. If the claims hold, the contribution is valuable: it provides a structured, machine-readable, queryable catalog of VA system components with explicit dependencies, and it demonstrates a semi-automated pipeline for scaling such knowledge bases. The paper's strengths include a public artifact, a clearly documented schema, a saturation-based justification for the initial coding set, and an honest discussion of limitations, including LLM granularity issues and the future need for multimodal input. The expert interviews offer ecologically valid feedback from system builders. However, the current evidence does not yet support the temporal-evolution and completeness claims as strongly as the paper suggests, because of the mixed manual/LLM extraction pipeline and the self-referential evaluation design. With targeted robustness analyses and more independent validation, the contribution would be solid.

major comments (4)
  1. [§3.4.2, §3.5, §5.3, §6.1, Table 1] The knowledge base is built from two very different extraction protocols: 20 systems manually coded from both text and figures, and 81 systems extracted by a text-only LLM. The authors acknowledge in Section 8 that the current pipeline 'relies on textual content' and plans multimodal input. The expert interviews in Section 6.1 found missing data sources, processing components, and feedback edges of exactly the type that are often conveyed in figures, and Table 1 shows LLM under-extraction concentrated in Data Processing (-22%) and Visualization (-15%). This is a correctness risk for the global statistics in Section 5.1 and, more importantly, for the temporal and diversity trends in Section 5.3, because the fraction of manually versus LLM-extracted systems may differ across periods. The authors should stratify the Section 5.3 analysis by extraction method, or re-run it on LLM-extracted sy
  2. [§6.2, Table 1] The quantitative evaluation compares LLM outputs to manual annotations created by the same team that designed the schema, codebook, and few-shot examples. This creates a circularity risk: the comparison largely measures how well the LLM reproduces the team's interpretive decisions, not how well either captures the systems. The paper does not state whether the annotators of the 20 evaluation papers were blind to the LLM outputs or whether a reliability study was conducted on those manual annotations. To support the claimed 86% label-match rate and the under-extraction findings, the authors should add an independent inter-annotator study, or at minimum report agreement between the evaluation annotations and the original 20-paper coding, and discuss how the results would change with independent ground truth.
  3. [§3.4.2] The manual open-coding process is described as one graduate assistant per paper with weekly team discussions, but no inter-annotator reliability metric is reported. Since this manual set defines the codebook, provides the few-shot examples, and serves as ground truth for Section 6.2, the consistency of the coding is load-bearing. The saturation claim ('relative rate of new label discovery fell below 3% by the 20th paper') also depends on the chosen threshold and on the ordering of the 20 papers. I would like to see a reliability statistic on a subset of doubly coded papers, a description of how the 20 papers were ordered and how the 3% threshold was set, and a sensitivity check showing that the core schema labels would not change materially under a different stopping rule.
  4. [§3.3] The corpus curation is not fully reproducible. The 101 papers are drawn from the authors' prior surveys [16,37] plus manual review, but the paper provides no inclusion/exclusion log, no PRISMA-style flow diagram, no list of venues and search dates, and no full corpus list in the main text. Because all downstream frequency and temporal analyses (Section 5) describe this corpus, a reader cannot assess how selection criteria influence the findings. The authors should provide the full paper list and screening decisions, either in an appendix or in the public repository, and state explicitly whether any papers were excluded after full-text review and for what reasons.
minor comments (5)
  1. [Table 1] The 'Difference' column is described in the text as 'mean absolute difference across the 20 papers,' but the table header simply says 'Difference.' Please rename to 'Mean Abs. Diff.' or similar to avoid confusion with the difference of the two means.
  2. [§3.5.1] For LLM reproducibility, please report the exact model version (e.g., GPT-4 snapshot/date), temperature, max tokens, and provide the full prompt template in the repository. The current description ('GPT-4' with few-shot prompting) is not sufficient for replication.
  3. [Throughout] There are inconsistent typographic variants of the paper's central term: 'VA' and 'V A' appear interchangeably. Please use one form consistently.
  4. [§5.3] The temporal trends are reported descriptively without confidence intervals or significance tests. Given the small per-year sample sizes (e.g., 9 systems in 2024), adding uncertainty bounds or at least a per-year error bar would make the claims easier to evaluate.
  5. [References] Reference [62] has a malformed DOI ('arXiv.2302.080811' is missing a digit; the correct arXiv number appears to be 2302.08081). Please correct it.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: knowledge base content is derived from external research papers and expert validation; self-citations are not load-bearing.

full rationale

The derivation chain (Section 3) is: curate 101 external papers -> manually code an initial 20 (Section 3.4.2) -> define a JSON schema -> use GPT-4 with few-shot examples to extract the remaining 81 (Section 3.5) -> validate via expert interviews (Section 6.1) and a held-out quantitative comparison (Section 6.2). Each link depends on the content of the source papers, not on the paper's own conclusions. The only self-citations are the use of the authors' prior urban-VA surveys [16,37] to seed the corpus (Section 3.3); these are independent published reviews and do not define the target component vocabulary, so they are not load-bearing circularity. The LLM evaluation uses 20 papers 'not included in our initial dataset' with fresh manual annotations as ground truth, so it is not a fitted-input-called-prediction. The stated limitations (Section 8: LLM 'currently relies on textual content' and future multimodal work; Section 6.1: experts found 7 missing granular blocks and 5 missing edges) are external-validity or completeness concerns, not equivalence-by-construction. Even if figure-only components are under-extracted, that makes the catalog incomplete, not circular. No equation or definition makes the knowledge base equivalent to its inputs. Score 0.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The paper introduces a multi-level classification schema (high-level, intermediate, granular blocks) and dependency types, but these are conceptual categories, not new physical or technological entities. The only hand-picked parameters are the saturation threshold, the temporal split year, and the initial coding set size; none are fitted to data in the numerical sense, but all influence the reported results.

free parameters (3)
  • Saturation threshold for manual coding = 3% new labels per paper
    The authors stopped expanding the manual set after the relative rate of new component-label discovery fell below 3% (Section 3.4.2). The 3% threshold and the stopping rule are chosen by hand; they determine the 20-paper seed set and thus the initial component vocabulary.
  • Temporal split year for evolution analysis = 2017
    In Section 5.3, component diversity before and after 2017 is compared (2007-2016 vs 2017-2024). The split is arbitrary and affects the reported 140-260% diversity increases.
  • Initial manual coding set size = 20 papers
    The schema and codebook are derived from an initial set of 20 papers (Section 3.4.2). This size is justified by the saturation rule but is otherwise a design choice; a different seed set could yield a different taxonomy.
assumptions (4)
  • domain assumption Research papers are accurate, complete, and unambiguous descriptions of the VA systems they present.
    The entire extraction pipeline reads papers, not source code, as stated in Section 2.1 and 3.5.1. Expert interviews found missing components, so this assumption is imperfect.
  • domain assumption All VA systems can be adequately represented as directed dataflow graphs of components and operations.
    Section 3.4.1 asserts that even systems that do not expose a dataflow implicitly follow one. This justifies the schema but may not capture control-flow or stateful behavior.
  • domain assumption GPT-4 can follow the provided JSON schema to extract structured components from full papers.
    Section 3.5.1 relies on GPT-4's few-shot structured extraction abilities without comparing models or reporting temperature/decoding settings.
  • domain assumption Manual coding by three researchers with weekly meetings produces a reliable gold standard.
    Sections 3.4.2 and 6.2 use manually coded labels as ground truth without inter-annotator reliability statistics (e.g., Cohen's kappa).

how reviews work

0 comments
Cite this review

Pith. "Pith review of VA-Blueprint: Uncovering Building Blocks for Visual Analytics System Design." pith.science (2026). https://pith.science/paper/XYUICOO7

@misc{pith2026250807497,
  author       = {Pith},
  title        = {Pith review of: VA-Blueprint: Uncovering Building Blocks for Visual Analytics System Design},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/XYUICOO7}},
  note         = {Machine review of arXiv:2508.07497}
}
read the original abstract

Designing and building visual analytics (VA) systems is a complex, iterative process that requires the seamless integration of data processing, analytics capabilities, and visualization techniques. While prior research has extensively examined the social and collaborative aspects of VA system authoring, the practical challenges of developing these systems remain underexplored. As a result, despite the growing number of VA systems, there are only a few structured knowledge bases to guide their design and development. To tackle this gap, we propose VA-Blueprint, a methodology and knowledge base that systematically reviews and categorizes the fundamental building blocks of urban VA systems, a domain particularly rich and representative due to its intricate data and unique problem sets. Applying this methodology to an initial set of 20 systems, we identify and organize their core components into a multi-level structure, forming an initial knowledge base with a structured blueprint for VA system development. To scale this effort, we leverage a large language model to automate the extraction of these components for other 81 papers (completing a corpus of 101 papers), assessing its effectiveness in scaling knowledge base construction. We evaluate our method through interviews with experts and a quantitative analysis of annotation metrics. Our contributions provide a deeper understanding of VA systems' composition and establish a practical foundation to support more structured, reproducible, and efficient system development. VA-Blueprint is available at https://urbantk.org/va-blueprint.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

73 extracted references · 32 canonical work pages

  1. [1]

    Agarwal, G

    S. Agarwal, G. Sahu, A. Puri, I. H. Laradji, K. D. Dvijotham, J. Stanley, L. Charlin, and C. Pal. LitLLMs, LLMs for literature review: Are we there yet?, 2025. doi: 10.48550/arXiv.2412.15249 2

  2. [2]

    Akbaba, D

    D. Akbaba, D. Lange, M. Correll, A. Lex, and M. Meyer. Troubling Col- laboration: Matters of care for visualization design study. InProceedings of the 2023 CHI Conference on Human Factors in Computing Systems, CHI ’23, article no. 812, 15 pages. Association for Computing Machinery, New York, NY , USA, 2023. doi: 10.1145/3544548.3581168 2

  3. [3]

    Al-Dohuki, Y

    S. Al-Dohuki, Y . Wu, F. Kamw, J. Yang, X. Li, Y . Zhao, X. Ye, W. Chen, C. Ma, and F. Wang. SemanticTraj: A new approach to interacting with massive taxi trajectories.IEEE Transactions on Visualization and Com- puter Graphics, 23(1):11–20, 2017. doi: 10.1109/TVCG.2016.2598416 6

  4. [4]

    Boorboor, Y

    S. Boorboor, Y . Kim, P. Hu, J. M. Moses, B. A. Colle, and A. E. Kaufman. Submerse: Visualizing storm surge flooding simulations in immersive display ecologies.IEEE Transactions on Visualization and Computer Graphics, 30(9):6365–6377, 2024. doi: 10.1109/TVCG.2023.3332511 4, 7

  5. [5]

    Brehmer and T

    M. Brehmer and T. Munzner. A multi-level typology of abstract visualiza- tion tasks.IEEE Transactions on Visualization and Computer Graphics, 19(12):2376–2385, 2013. doi: 10.1109/TVCG.2013.124 1, 2

  6. [6]

    Cao, Y .-R

    N. Cao, Y .-R. Lin, X. Sun, D. Lazer, S. Liu, and H. Qu. Whisper: Tracing the spatiotemporal process of information diffusion in real time.IEEE Transactions on Visualization and Computer Graphics, 18(12):2649–2658,

  7. [7]

    J. H. Caufield, H. Hegde, V . Emonet, N. L. Harris, M. P. Joachimiak, N. Matentzoglu, H. Kim, S. Moxon, J. T. Reese, M. A. Haendel, P. N. Robinson, and C. J. Mungall. Structured prompt interrogation and recur- sive extraction of semantics (SPIRES): A method for populating knowl- edge bases using zero-shot learning.Bioinformatics, 40(3):btae104, 02

  8. [8]

    J. Chen, Q. Huang, C. Wang, and C. Li. SenseMap: Urban performance visualization and analytics via semantic textual similarity.IEEE Transac- tions on Visualization and Computer Graphics, 30(9):6275–6290, 2024. doi: 10.1109/TVCG.2023.3333356 4

Show all 73 references
  1. [9]

    Chen and D

    M. Chen and D. S. Ebert. An ontological framework for supporting the design and evaluation of visual analytics systems.Computer Graphics Forum, 38(3):131–144, 2019. doi: 10.1111/cgf.13677 2

  2. [10]

    M. Chen, G. Grinstein, C. R. Johnson, J. Kennedy, and M. Tory. Pathways for theoretical advances in visualization.IEEE Computer Graphics and Applications, 37(4):103–112, 2017. doi: 10.1109/MCG.2017.3271463 2

  3. [11]

    W. Chen, Z. Huang, F. Wu, M. Zhu, H. Guan, and R. Maciejewski. V AUD: A visual analysis approach for exploring spatio-temporal urban data.IEEE Transactions on Visualization and Computer Graphics, 24(9):2636–2648,

  4. [12]

    W. Cui. Visual Analytics: A comprehensive overview.IEEE Access, 7:81555–81573, 2019. doi: 10.1109/ACCESS.2019.2923736 1

  5. [14]

    D. Duke, K. Brodlie, and D. Duce. Building an ontology of visualization. InIEEE Visualization 2004, pp. 7p–7p, 2004. doi: 10.1109/VISUAL.2004 .10 2

  6. [15]

    D. Duke, K. Brodlie, D. Duce, and I. Herman. Do you see what i mean? [data visualization].IEEE Computer Graphics and Applications, 25(3):6– 9, 2005. doi: 10.1109/MCG.2005.55 2

  7. [16]

    Ferreira, G

    L. Ferreira, G. Moreira, M. Hosseini, M. Lage, N. Ferreira, and F. Miranda. Assessing the landscape of toolkits, frameworks, and authoring tools for urban visual analytics systems.Computers & Graphics, 123:104013, 2024. doi: 10.1016/j.cag.2024.104013 2, 3

  8. [17]

    Ferreira, M

    N. Ferreira, M. Lage, H. Doraiswamy, H. V o, L. Wilson, H. Werner, M. Park, and C. Silva. Urbane: A 3D framework to support data driven decision making in urban development. In2015 IEEE Conference on Visual Analytics Science and Technology (VAST), pp. 97–104, 2015. doi: 10.110...

  9. [18]

    Ferreira, J

    N. Ferreira, J. Poco, H. T. V o, J. Freire, and C. T. Silva. Visual exploration of big spatio-temporal urban data: A study of New York City taxi trips. IEEE Transactions on Visualization and Computer Graphics, 19(12):2149– 2158, 2013. doi: 10.1109/TVCG.2013.226 7

  10. [19]

    Garcia, J

    G. Garcia, J. Silveira, J. Poco, A. Paiva, M. B. Nery, C. T. Silva, S. Adorno, and L. G. Nonato. CrimAnalyzer: Understanding crime patterns in São Paulo.IEEE Transactions on Visualization and Computer Graphics, 27(4):2313–2328, 2021. doi: 10.1109/TVCG.2019.2947515 4

  11. [20]

    Gautier, M

    J. Gautier, M. Brédif, and S. Christophe. Co-visualization of air temper- ature and urban data for visual exploration. In2020 IEEE Visualization Conference (VIS), pp. 71–75, 2020. doi: 10.1109/VIS47514.2020.00021 1

  12. [21]

    Gilson, N

    O. Gilson, N. Silva, P. Grant, and M. Chen. From web data to visualization via ontology mapping.Computer Graphics Forum, 27(3):959–966, 2008. doi: 10.1111/j.1467-8659.2008.01230.x 2

  13. [22]

    Heer and M

    J. Heer and M. Agrawala. Software design patterns for information visu- alization.IEEE Transactions on Visualization and Computer Graphics, 12(5):853–860, 2006. doi: 10.1109/TVCG.2006.178 4

  14. [23]

    Hoque and M

    E. Hoque and M. Agrawala. Searching the visual style and structure of d3 visualizations.IEEE Transactions on Visualization and Computer Graphics, 26(1):1236–1245, 2020. doi: 10.1109/TVCG.2019.2934431 2

  15. [24]

    Isenberg

    T. Isenberg. The state of reproducibility stamps for visualization research papers. In2024 IEEE Evaluation and Beyond - Methodological Ap- proaches for Visualization (BELIV), pp. 97–105. IEEE Computer Society, Los Alamitos, CA, USA, 2024. doi: 10.1109/BELIV64461.2024.00016 2

  16. [25]

    M. Kay. ggdist: Visualizations of distributions and uncertainty in the grammar of graphics.IEEE Transactions on Visualization and Computer Graphics, 30(1):414–424, 2024. doi: 10.1109/TVCG.2023.3327195 2

  17. [26]

    R. M. Kirby and M. Meyer. Visualization Collaborations: What works and why.IEEE Computer Graphics and Applications, 33(6):82–88, 2013. doi: 10.1109/MCG.2013.101 2

  18. [27]

    Krueger, J

    R. Krueger, J. Beyer, W.-D. Jang, N. W. Kim, A. Sokolov, P. K. Sorger, and H. Pfister. Facetto: Combining unsupervised and supervised learning for hierarchical phenotype analysis in multi-channel image data.IEEE Transactions on Visualization and Computer Graphics, 26(1):227–237,

  19. [29]

    H. Li, Y . Wang, A. Wu, H. Wei, and H. Qu. Structure-aware visualization retrieval. InProceedings of the 2022 CHI Conference on Human Factors in Computing Systems, CHI ’22, article no. 409, 14 pages. Association for Computing Machinery, New York, NY , USA, 2022. doi: 10.1145/3...

  20. [30]

    P. Li, X. Jiang, G. Zhang, J. T. Trabucco, D. Raciti, C. Smith, M. Ring- wald, G. E. Marai, C. Arighi, and H. Shatkay. Utilizing image and cap- tion information for biomedical document classification.Bioinformatics, 37:i468–i476, 07 2021. doi: 10.1093/bioinformatics/btab331 2

  21. [31]

    D. Liu, P. Xu, and L. Ren. TPFlow: Progressive partition and multidi- mensional pattern extraction for large-scale spatio-temporal data analysis. IEEE Transactions on Visualization and Computer Graphics, 25(1):1–11,

  22. [32]

    S. LYi, Q. Wang, F. Lekschas, and N. Gehlenborg. Gosling: A grammar- based toolkit for scalable and interactive genomics data visualization.IEEE Transactions on Visualization and Computer Graphics, 28(1):140–150,

  23. [33]

    Y . Lyu, H. Lu, M. K. Lee, G. Schmitt, and B. Y . Lim. IF-City: Intelligible fair city planning to measure, explain and mitigate inequality.IEEE Transactions on Visualization and Computer Graphics, 30(7):3749–3766,

  24. [34]

    A. M. McNutt. No Grammar to Rule Them All: A survey of JSON-style DSLs for visualization.IEEE Transactions on Visualization and Computer Graphics, 29(1):160–170, 2023. doi: 10.1109/TVCG.2022.3209460 2

  25. [35]

    Meyer and J

    M. Meyer and J. Dykes. Criteria for rigor in visualization design study. IEEE Transactions on Visualization and Computer Graphics, 26(1):87–97,

  26. [36]

    Minaee, T

    S. Minaee, T. Mikolov, N. Nikzad, M. Chenaghlu, R. Socher, X. Amatriain, and J. Gao. Large language models: A survey, 2025. doi: 10.48550/arXiv. 2402.06196 2

  27. [37]

    Miranda, T

    F. Miranda, T. Ortner, G. Moreira, M. Hosseini, M. Vuckovic, F. Biljecki, C. T. Silva, M. Lage, and N. Ferreira. The state of the art in visual analytics for 3D urban data.Computer Graphics Forum, 43(3):e15112, 2024. doi: 10.1111/cgf.15112 3

  28. [38]

    Moreira, M

    G. Moreira, M. Hosseini, M. N. Alam Nipu, M. Lage, N. Ferreira, and F. Miranda. The Urban Toolkit: A grammar-based framework for ur- ban visual analytics.IEEE Transactions on Visualization and Computer Graphics, 30(1):1402–1412, 2024. doi: 10.1109/TVCG.2023.3326598 2

  29. [39]

    Moreira, M

    G. Moreira, M. Hosseini, C. Veiga, L. Alexandre, N. Colaninno, D. de Oliveira, N. Ferreira, M. Lage, and F. Miranda. Curio: A dataflow- based framework for collaborative urban visual analytics.IEEE Transac- tions on Visualization and Computer Graphics, 31(1):1224–1234, 2025. d...

  30. [40]

    doi: 10.1109/TVCG.2023.3239909 4, 7

  31. [41]

    Nonnemann, H

    L. Nonnemann, H. Schumann, B. Urban, M. Aehnelt, and H.-J. Schulz. A characterization of data exchange between visual analytics tools. In 2020 24th International Conference Information Visualisation (IV), pp. 368–377, 2020. doi: 10.1109/IV51561.2020.00066 4

  32. [42]

    A. M. Pérez, C. P. Risquet, and J. M. Gómez. An enhanced visualization ontology for a better representation of the visualization process. In M. Gu- sev and P. Mitrevski, eds.,ICT Innovations 2010, pp. 342–347. Springer, Berlin, Heidelberg, 2011. doi: 10.1007/978-3-642-19325-5_35 2

  33. [43]

    doi: 10.1109/TVCG.2019.2934539 2

  34. [44]

    Polowinski and M

    J. Polowinski and M. V oigt. VISO: A shared, formal knowledge base as a foundation for semi-automatic infovis systems. InCHI ’13 Extended Abstracts on Human Factors in Computing Systems, CHI EA ’13, 6 pages, p. 1791–1796. Association for Computing Machinery, New York, NY , USA...

  35. [45]

    X. Pu, M. Gao, and X. Wan. Summarization is (almost) dead, 2023. doi: 10.48550/arXiv.2309.09558 2

  36. [46]

    Rauer-Zechmeister, D

    S. Rauer-Zechmeister, D. Cornel, B. Sadransky, Z. Horváth, A. Konev, A. Buttinger-Kreuzhuber, R. Heidrich, G. Blöschl, E. Gröller, and J. Waser. HORA 3D: Personalized flood risk visualization as an interactive web service.Computer Graphics Forum, 43(3):e15110, 2024. doi: 10.11...

  37. [47]

    Rulff, F

    J. Rulff, F. Miranda, M. Hosseini, M. Lage, M. Cartwright, G. Dove, J. Bello, and C. T. Silva. Urban Rhapsody: Large-scale exploration of urban soundscapes.Computer Graphics Forum, 41(3):209–221, 2022. doi: 10.1111/cgf.14534 6

  38. [48]

    Moritz, C

    D. Moritz, C. Wang, G. L. Nelson, H. Lin, A. M. Smith, B. Howe, and J. Heer. Formalizing Visualization Design Knowledge as Constraints: Ac- tionable and extensible models in draco.IEEE Transactions on Visualiza- tion and Computer Graphics, 25(1):438–448, 2019. doi: 10.1109/TVC...

  39. [50]

    Sedlmair, M

    M. Sedlmair, M. Meyer, and T. Munzner. Design Study Methodology: Reflections from the trenches and the stacks.IEEE Transactions on Visualization and Computer Graphics, 18(12):2431–2440, 2012. doi: 10. 1109/TVCG.2012.213 2, 4

  40. [51]

    Poco and J

    J. Poco and J. Heer. Reverse-engineering visualizations: Recovering visual encodings from chart images.Computer Graphics Forum, 36(3):353–363,

  41. [52]

    Sicat, J

    R. Sicat, J. Li, J. Choi, M. Cordeil, W.-K. Jeong, B. Bach, and H. Pfister. DXR: A toolkit for building immersive data visualizations.IEEE Transac- tions on Visualization and Computer Graphics, 25(1):715–725, 2019. doi: 10.1109/TVCG.2018.2865152 2

  42. [53]

    X. Tang, E. Krokos, C. Liu, K. Davidson, K. Whitley, N. Ramakrishnan, and C. North. Steering LLM summarization with visual workspaces for sensemaking, 2024. doi: 10.48550/arXiv.2409.17289 2

  43. [54]

    Wagner, C

    J. Wagner, C. T. Silva, W. Stuerzlinger, and L. Nedel. Reimagining TaxiVis through an immersive space-time cube metaphor and reflecting on potential benefits of immersive analytics for urban data exploration. In 2024 IEEE Conference Virtual Reality and 3D User Interfaces (VR),...

  44. [55]

    G. H. Weber, S. Carpendale, D. Ebert, B. Fisher, H. Hagen, B. Shnei- derman, and A. Ynnerman. Apply or Die: On the role and assessment of application papers in visualization.IEEE Computer Graphics and Applications, 37(3):96–104, 2017. doi: 10.1109/MCG.2017.51 2

  45. [56]

    C. Weil, S. E. Bibri, R. Longchamp, F. Golay, and A. Alahi. Urban digital twin challenges: A systematic review and perspectives for sustainable smart cities.Sustainable Cities and Society, 99:104862, 2023. doi: 10. 1016/j.scs.2023.104862 2

  46. [57]

    Sacha, M

    D. Sacha, M. Kraus, D. A. Keim, and M. Chen. VIS4ML: An ontology for visual analytics assisted machine learning.IEEE Transactions on Visualization and Computer Graphics, 25(1):385–395, 2019. doi: 10. 1109/TVCG.2018.2864838 2

  47. [58]

    A. Wu, D. Deng, M. Chen, S. Liu, D. Keim, R. Maciejewski, S. Miksch, H. Strobelt, F. Viégas, and M. Wattenberg. Grand challenges in visual ana- lytics applications.IEEE Computer Graphics and Applications, 43(5):83– 90, 2023. doi: 10.1109/MCG.2023.3284620 1, 2

  48. [59]

    A. Wu, D. Deng, F. Cheng, Y . Wu, S. Liu, and H. Qu. In Defence of Visual Analytics Systems: Replies to Critics .IEEE Transactions on Visualization & Computer Graphics, 29(01):1026–1036, jan 2023. doi: 10. 1109/TVCG.2022.3209360 2

  49. [60]

    B. Selic. The pragmatics of model-driven development.IEEE Software, 20(5):19–25, 2003. doi: 10.1109/MS.2003.1231146 2

  50. [61]

    C. Yang, Z. Zhang, Z. Fan, R. Jiang, Q. Chen, X. Song, and R. Shibasaki. EpiMob: Interactive visual analytics of citywide human mobility restric- tions for epidemic control.IEEE Transactions on Visualization and Computer Graphics, 29(8):3586–3601, 2023. doi: 10.1109/TVCG.2022....

  51. [62]

    X. Yang, Y . Li, X. Zhang, H. Chen, and W. Cheng. Exploring the limits of chatgpt for query or aspect-based text summarization, 2023. doi: 10. 48550/arXiv.2302.080811 2

  52. [63]

    L. Ying, Y . Wang, H. Li, S. Dou, H. Zhang, X. Jiang, H. Qu, and Y . Wu. Reviving static charts into live charts.IEEE Transactions on Visualiza- tion and Computer Graphics, pp. 1–16, 2024. doi: 10.1109/TVCG.2024. 3397004 2

  53. [64]

    L. Ying, A. Wu, H. Li, Z. Deng, J. Lan, J. Wu, Y . Wang, H. Qu, D. Deng, and Y . Wu. V AID: Indexing view designs in visual analytics system. In Proceedings of the 2024 CHI Conference on Human Factors in Computing Systems, CHI ’24, article no. 198, 15 pages. Association for Co...

  54. [65]

    Yu and C

    B. Yu and C. T. Silva. VisFlow - web-based visualization framework for tabular data with a subset flow model.IEEE Transactions on Visualization and Computer Graphics, 23(1):251–260, 2017. doi: 10.1109/TVCG.2016. 2598497 4

  55. [66]

    D. Weng, C. Zheng, Z. Deng, M. Ma, J. Bao, Y . Zheng, M. Xu, and Y . Wu. Towards better bus networks: A visual analytics approach.IEEE Transactions on Visualization and Computer Graphics, 27(2):817–827,

  56. [67]

    Y . Zhou, X. Meng, Y . Wu, T. Tang, Y . Wang, and Y . Wu. An intelligent approach to automatically discovering visual insights.Journal of Visu- alization, 26(3):705–722, June 2023. doi: 10.1007/s12650-022-00894-z 2

  57. [70]

    W. Wu, Y . Zheng, N. Cao, H. Zeng, B. Ni, H. Qu, and L. M. Ni. MobiSeg: Interactive region segmentation using heterogeneous mobility data. In 2017 IEEE Pacific Visualization Symposium (PacificVis), pp. 91–100, 2017. doi: 10.1109/PACIFICVIS.2017.8031583 6

  58. [76]

    Zhang, P

    H. Zhang, P. S. Yu, and J. Zhang. A systematic survey of text summariza- tion: From statistical methods to large language models, 2024. doi: 10. 48550/arXiv.2406.11289 2

  59. [2012]

    doi: 10.1109/TVCG.2012.291 4

  60. [2017]

    doi: 10.1111/cgf.13193 2

  61. [2018]

    doi: 10.1109/TVCG.2017.2758362 4, 6, 7

  62. [2019]

    doi: 10.1109/TVCG.2018.2865018 6, 7

  63. [2021]

    doi: 10.1109/TVCG.2020.3030458 4

  64. [2022]

    doi: 10.1109/TVCG.2021.3114876 2

  65. [2024]

    doi: 10.1093/bioinformatics/btae104 5

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.