Two comparisons, each split by task. Table 1 (C3 vs. C4) shows, for every library paper either design cited, what kind of information was drawn from it — empty where a design didn't use that paper at all. Table 2 (C4 vs. C5) shows, for every distinct method named across the four designs, which of them included it. C5 continues C3 (not C4), so its source-paper set differs from C4's own citations by design — the point of Table 2 is exactly to show how much methodological ground the two independent paths (C4's own citations vs. C3's citations enriched in C5) end up covering in common.
Table 1 — Library papers used: C3 vs. C4
VAST
| Library paper | C3 / Sonnet | C3 / Terra | C4 / Sonnet | C4 / Terra |
|---|---|---|---|---|
| Andrienko et al. (2011) — From Movement Tracks through Events to Places | Spatio-temporal aggregation + two-stage density clustering, adapted for zone-level incidence aggregation and fine-grained origin localisation. | Iterative extraction, clustering, visual inspection, and refinement of spatial concentrations — adapted for early illness-report clusters and candidate origin areas. | Core event-to-place extraction logic (Stage C): density-clustered “significant places” via combined spatio-temporal distance functions, aggregated into per-place time series. | Density-based clustering of repeated spatial/temporal events to identify significant locations; adapted from movement events to symptom-report hotspots. |
| Choo et al. (2013) — UTOPIAN (Human-Steered Topic Modelling) | — | — | — | Iteratively classify and refine microblog topics with analyst input; inspect documents, correct misleading terms, merge/split themes, and retrain. |
| Andrienko & Andrienko (2013) — A Visual Analytics Framework for Spatio-Temporal Analysis and Modelling | — | Main iterative cycle: spatial-temporal aggregation, grouping/modelling, visual fit assessment, residual analysis, model refinement. | Zone-grouping, representative-curve modelling, and residual-inspection logic (Stage D): cluster the 13 zones by temporal incidence similarity, check spatial coherence. | Aggregate into spatial time series; group areas with similar temporal behaviour; model temporal patterns; inspect residuals to refine explanations. |
| Andrienko et al. (2024) — Human-in-the-Loop: Visual Analytics for Building Models Recognizing Behavioral Patterns | Iterative criteria/feature-refinement loop pattern underlying the extraction-validation loop. | — | Human-in-the-loop validation principle (Stage A): visually check extracted symptom/severity indicators before trusting them for further analysis. | — |
| Wyss et al. (2025) — DaedalusData | Interactive labelling/refinement loop pattern informing the extraction-criteria refinement loop. | — | — | — |
| Andrienko et al. (2026) — Detecting Stable Cross-Impact Patterns in Bivariate Time Series | Lagged cross-impact detection methodology, adapted for testing wind vs. incidence correlation under the airborne-transmission hypothesis. | Lagged relationships, parameter calibration, sensitivity analysis, comparison of directional cross-effects (e.g. wind-aligned or delayed propagation). | Lagged cross-impact/Ikat-plot methodology (Stage E): test whether wind direction/timing precedes consistent downwind incidence increases. | Examine lagged relationships between a possible driver and later outcomes; suggested for wind conditions vs. downwind symptom-report activity. |
| Bernard et al. (2025) — IVESA | Metadata relation-seeking multi-view design pattern underlying the multi-view hypothesis-testing arrangement (Phase 3). | — | — | — |
| Sachdeva et al. (2023) — LFPeers | Correlation-exploration loop pattern (query→compute→visualise→assess→refine) underlying the iterative hypothesis-testing loop. | — | — | — |
| Kurzhals, Heimerl & Weiskopf (2014) — ISeeCube | — | Coordinated exploration design: linked spatial, temporal, and space-time views used to identify patterns before detailed interpretation. | — | — |
| Stasko, Görg & Liu (2008) — Jigsaw | — | — | Entity/keyword extraction from microblog text (Stage A); human correction of errors; raw-text drill-down/document verification (Stage G); co-occurrence/relation-extraction for person-to-person testing (Stage F). | Extract and validate entities (places, hospitals, landmarks, organizations, dates) from posts; keep links to original text for evidential inspection. |
| van Wijk & van Selow (1999) — Cluster and Calendar Based Visualization | — | — | Daily-profile clustering and baseline-deviation detection (Stage B): detect the onset day(s) breaking from the pre-outbreak pattern. | — |
| Mühlbacher & Piringer (2013) — Partition-Based Regression Modelling | — | — | — | Build interpretable explanatory models incrementally, inspect feature relationships visually, then use residual patterns to identify missing factors/interactions. |
| von Landesberger et al. (2016) — MobilityGraphs | — | — | — | Suggested as an optional extension if anonymised aggregate mobility flows become available, to assess whether mobility explains outbreak progression. |
VIS
| Library paper | C3 / Sonnet | C3 / Terra | C4 / Sonnet | C4 / Terra |
|---|---|---|---|---|
| van Wijk & van Selow (1999) — Cluster and Calendar Based Visualization | Model for the calibration loop with visual assessment of clustering resolution parameters. | Analyst-guided iterative refinement pattern: inspect clustering results, assess interpretability, adjust parameters, repeat. | — | — |
| von Landesberger et al. (2016) — MobilityGraphs | Model for the parameter-refinement loop pattern (quality heatmap→assess→adjust) applied to overlap-threshold calibration. | — | Source of the “full-period supergraph as anchor, then per-window analysis” pattern used to keep temporal simplification consistent. | — |
| Guo et al. (2018) — EventThread | Model for tensor-decomposition-into-threads-with-branching/merging, adapted to community-lineage tracking (continuation/merge/split/birth/dissolution). | Inspiration for representing community evolution as temporal threads/lineages: branching, convergence, splitting, merging, emergence, disappearance. | Source of the tensor-decomposition “latent thread” concept adapted for robust community-trajectory tracking beyond set-overlap matching. | — |
| Streit et al. (2014) — StratomeX | Model for the column-band arrangement visualising overlapping set memberships across adjacent groupings. | — | Source of the overlap-band/query-based comparison logic adapted for comparing community memberships across time periods. | Used for comparing overlapping groups and tracing membership across alternative or evolving groupings. |
| Henry, Fekete & McGuffin (2007) — NodeTrix | Model for the hybrid node-link/matrix representation used to inspect dense community instances and identify actor roles (leaders, bridges). | Used for the proposed hybrid community-detail view: dense membership as matrices, inter-community ties/bridging authors as node-link. | Directly informs the hybrid node-link/matrix inspection of individual communities' internal structure (core-periphery, bridging authors). | Used for detailed inspection of dense collaboration communities while preserving links between communities. |
| Kwon et al. (2021) — DPVis | Model for state-persistence visualisation and role/leadership identification via matrix-based presence tracking over time. | — | — | — |
| Choo et al. (2013) — UTOPIAN | Model for the interactive topic-modelling refinement loop used to derive and iteratively adjust research-focus labels. | Used for the iterative topic-modelling component: inspect topic coherence, refine definitions/parameters, interpret as research foci. | Source of the interactive NMF-based topic modeling approach (merge/split/seed topics) for extracting/refining research foci. | Used for characterizing communities' research foci from titles/abstracts/keywords; analyst-guided topic refinement. |
| Andrienko & Andrienko (2013) — A Visual Analytics Framework for Spatio-Temporal Analysis and Modelling | Model for the nested-loop structure combining iterative clustering refinement with parameter calibration. | — | — | Relevant to iterative clustering of temporal profiles and analyst validation of cluster coherence. |
| Bernard et al. (2025) — IVESA | Model for the exploratory synthesis loop with implicit termination criterion (“sufficient holistic understanding”). | — | Additional source for the overview-then-detail sequence timeline pattern used for author-level membership tracks. | — |
| Bernard et al. (2013) — MotionExplorer | Additional reference for exploratory loops with implicit, judgment-based termination conditions. | — | — | — |
| van den Elzen et al. (2016) — Reducing Snapshots to Points | — | Used for constructing and comparing temporal co-authorship-network snapshots, detecting stable states and transitions, and using projection-based exploration of network evolution. | Source of the windowed-snapshot construction and projection-based approach for visualizing network states as points/trajectories. | Used for representing time-windowed co-authorship networks as evolving network states; snapshot-state projection for stable periods/transitions. |
| Sacha et al. (2018) — SOMFlow | — | — | Source of the iterative “propose→inspect→adapt→partition→track provenance” loop applied to interface design/refinement. | Used for the iterative, human-guided approach to community detection and resolution selection; maintaining provenance of parameter choices. |
| Andrienko, Andrienko & Shirato (2023) — Progressive Abstraction Analysis (Episodes and Topics) | — | — | Source of the episode-to-topic abstraction pattern and matrix/small-multiples visualization for tracking topic evolution. | — |
| Sedlmair, Frank, Munzner & Butz (2012) — RelEx | — | — | Source of the matrix-based relational-pattern visualization adapted for the community-overlap matrix. | — |
| Monroe et al. (2013) — Temporal Event Sequence Simplification | — | — | Source of the aggregated event-sequence/scarf-plot idiom adapted for per-author community-membership timelines. | — |
| Spinner et al. (2020) — explAIner | — | — | Source of the general “understand→diagnose→refine” iterative loop structuring the overall analyst interaction cycle. | — |
| Pretorius & van Wijk (2006) — Visual Analysis of Multivariate State Transition Graphs | — | — | — | Relevant to analyzing transitions among abstract states; informs explicit treatment of continuations/splits/merges/births/dissolutions. |
| Stasko, Görg & Liu (2008) — Jigsaw | — | — | — | Relevant to grounding high-level patterns in original textual evidence; informs the paper-level inspector for validating interpretations. |
Table 2 — Methods included: C4 vs. C5
VAST
| Method (source paper) | C4 / Sonnet | C4 / Terra | C5 / Sonnet | C5 / Terra |
|---|---|---|---|---|
| Entity/keyword extraction from text (Jigsaw) | ✓ | ✓ | — | — |
| Topic modelling for message classification (UTOPIAN) | — | ✓ | — | — |
| HITL feature-engineering scaffold: UMAP + k-means/silhouette + k-NN validation (HITL) | ✓ | — | ✓ | — |
| Two-stage DBSCAN/OPTICS density clustering w/ multidimensional distance function (MovTracks / Places) | ✓ | ✓ | ✓ | ✓ |
| Visual/interactive threshold estimation via nearest-neighbour distance (MovTracks / Places) | — | — | ✓ | ✓ |
| Buffer/convex-hull construction + infrastructure-based validation (MovTracks) | — | — | ✓ | — |
| Daily-profile clustering + baseline-deviation onset detection (Cluster and Calendar) | ✓ | — | — | — |
| Zone-grouping by temporal-profile similarity + residual-based refinement (ST-Model) | ✓ | ✓ | — | ✓ |
| Population-normalisation / relative-value transformation (ST-Model) | — | — | — | ✓ |
| Representative-curve construction (mean/median/trimmed) + Holt-Winters (ST-Model) | — | — | — | ✓ |
| Partition-based incremental regression modelling (Mühlbacher & Piringer) | — | ✓ | — | — |
| Tolerance-adjusted Kendall's Tau for lagged cross-impact (CrossImpact) | ✓ | ✓ | ✓ | ✓ |
| Ikat-plot visualisation + pattern typology (CrossImpact) | ✓ | — | ✓ | ✓ |
| Event extraction via connected-region / flood-fill search (CrossImpact) | — | — | ✓ | ✓ |
| Parameter-sensitivity analysis for cross-impact thresholds (CrossImpact) | — | — | ✓ | ✓ |
| Logistic-regression modelling of impact/transmission evidence (CrossImpact) | — | — | ✓ | — |
| Timeline chart with similarity-based zone ordering, vs. space-time cube (CrossImpact) | — | — | ✓ | — |
| Multi-particle / bulk visual inspection, lasso + rectangle selection (DaedalusData) | — | — | ✓ | — |
| Metadata-view heatmap cross-cut with capped binning (IVESA) | — | — | ✓ | — |
| Filtering view with opacity-preview of excluded items (IVESA) | — | — | ✓ | — |
| Search-then-explore two-phase structure (LFPeers) | — | — | ✓ | — |
| Correlation-ranking + drill-down scatterplot (LFPeers) | — | — | ✓ | — |
| Coordinated linked map/time-graph views with multi-attribute filtering (ST-Model / ISeeCube) | — | — | — | ✓ |
| Mobility-flow analysis, optional extension (MobilityGraphs) | — | ✓ | — | — |
VIS
| Method (source paper) | C4 / Sonnet | C4 / Terra | C5 / Sonnet | C5 / Terra |
|---|---|---|---|---|
| Temporal windowing/segmentation policy fixed as pre-process (IVESA) | — | — | ✓ | — |
| Supergraph pre-aggregation for cross-window comparability (MobilityGraphs) | ✓ | — | ✓ | — |
| Windowed/rolling network snapshots with overlap parameter (Reducing Snapshots to Points) | ✓ | ✓ | — | ✓ |
| Community detection: Louvain / DBSCAN-style with domain similarity criteria (MobilityGraphs) | — | — | ✓ | — |
| Parameter-grid heatmap for resolution calibration (MobilityGraphs) | — | — | ✓ | — |
| Top-down/bottom-up interactive cluster-count refinement (Cluster and Calendar) | — | — | ✓ | ✓ |
| NodeTrix hybrid node-link/matrix community view (NodeTrix) | ✓ | ✓ | ✓ | ✓ |
| Jaccard / relative-overlap similarity for lineage matching (StratomeX) | ✓ | ✓ | ✓ | — |
| Tensor decomposition (CP/PARAFAC) for joint thread/lineage extraction (EventThread) | ✓ | — | ✓ | — |
| Interactive threshold slider for lineage matching with visual feedback (EventThread) | — | — | ✓ | — |
| Stacked-block / alluvial lineage visualisation (StratomeX / EventThread) | — | — | ✓ | ✓ |
| Betweenness/degree centrality for core & bridge identification (NodeTrix) | — | — | ✓ | ✓ |
| Drag-and-drop removal test for bridge validation (NodeTrix) | — | — | ✓ | — |
| Feature-matrix heatmap + distribution view for author presence (DPVis) | — | — | ✓ | — |
| Pathway waterfall (beeswarm) + edge-bundling for trajectories (DPVis) | — | — | ✓ | — |
| Leadership-pattern typology: cross / block / intermediate (NodeTrix) | — | — | ✓ | — |
| TF-IDF term filtering/weighting (EventThread / UTOPIAN) | — | — | ✓ | ✓ |
| NMF topic modelling, preferred over LDA for run-to-run consistency (UTOPIAN) | ✓ | ✓ | ✓ | ✓ |
| Semi-supervised NMF interactive refinement: merge/split/seed topics (UTOPIAN) | ✓ | ✓ | ✓ | ✓ |
| Modified t-SNE with shrinkage parameter for topic-space visualisation (UTOPIAN) | — | — | ✓ | — |
| Motion-graph metaphor for author-trajectory overview (MotionExplorer) | — | — | ✓ | — |
| PCA-based deterministic layout for trajectory/state overview (MotionExplorer / Reducing Snapshots) | ✓ | ✓ | ✓ | ✓ |
| MDS/t-SNE alternative projections + distance-matrix comparison (Reducing Snapshots) | — | — | ✓ | ✓ |
| Pair-analytics collaborative interpretation protocol (DPVis) | — | — | ✓ | — |
| Graphical query builder / subgroup finder (DPVis) | — | — | ✓ | — |
| Residual-pattern inspection as loop-exit criterion (ST-Model) | — | ✓ | ✓ | — |
| External machine-readable model storage for reproducibility (ST-Model) | — | — | ✓ | — |
| State-transition/task-usage diagram reflecting on analyst workflow (IVESA) | ✓ | — | ✓ | — |
| DBLP-identifier-based author-disambiguation audit (NodeTrix) | — | — | — | ✓ |
| Event-sequence scarf-plot / aggregated timeline idiom (Monroe et al.) | ✓ | — | — | — |
| Matrix-based relational pattern for community-overlap (RelEx) | ✓ | — | — | — |
| Episode-to-topic abstraction + matrix/small-multiples (Progressive Abstraction Analysis) | ✓ | — | — | — |
| Understand-diagnose-refine iterative loop (explAIner) | ✓ | — | — | — |
| State-transition graph for community-evolution events (Pretorius & van Wijk) | — | ✓ | — | — |
| Text-evidence grounding / paper-level inspector (Jigsaw) | — | ✓ | — | — |
| SOMFlow propose-inspect-adapt-partition-track-provenance loop (SOMFlow) | ✓ | ✓ | — | — |