VIS · GPT-5.6-Terra · C7

Confirmed. I have absorbed the workflow library and its range of approaches, including temporal/event analysis, movement and network workflows, cohort comparison, visual machine learning, clustering, simulation steering, text analysis, and multimodal interaction analysis.

I will use the library as a design resource and cite relevant workflows by name when proposing the workflow for your task. I have no questions at this stage.

Evaluating citation networks

I'm considering how to handle a citation network if available, looking at time distributions and topic drift. I really need to avoid assigning the research to a multi-community paper. Instead, I'll assign the paper's contributions based on the authors’ community memberships and their respective fractions. I wonder how the leaders within those communities can influence this as well. It's definitely a bit of a balancing act!

Recommended workflow: Dynamic, overlapping co-authorship communities with topical characterization

This workflow treats the VIS literature as an evolving collaboration network, while preserving the fact that researchers may belong to several communities. It combines ideas from NodeTrix, Reducing Snapshots to Points, Human-Steered Topic Modelling, SOMFlow, and StratomeX.

1. Audit and prepare the publication data

First, validate the paper–author records:

• Use DBLP author identifiers as the primary identity key; retain raw names for display and manual checking.
• Identify duplicate papers, incomplete author lists, missing abstracts/keywords, and unusually large author teams.
• Inspect yearly publication counts by venue, because the four venues have different histories and may create apparent community changes caused by coverage rather than genuine intellectual change.
• Normalize titles, abstracts, and keywords: remove generic terms such as visualization, visual analytics, and venue names; retain domain terms and methodological phrases.
• If “citations” means citation links between papers, construct a paper-citation network. If it means only citation counts, use those as impact indicators rather than network links.

Create two linked data structures:

1. A paper–author bipartite network, preserving full author teams.
2. A weighted co-authorship network, where authors are nodes and a collaboration edge is created for each co-authored paper.

To avoid large author teams dominating the network, fractionally weight each paper’s collaboration contribution. For a paper with (n) authors, distribute a total contribution of 1 across its author pairs rather than assigning every pair a full weight.

---

2. Define temporal views at more than one scale

Construct overlapping temporal snapshots, for example:

• annual snapshots for detailed inspection;
• rolling 3- or 5-year windows for more stable community detection;
• broader eras, such as 1990–1999, 2000–2009, 2010–2019, and 2020–2024, for high-level reporting.

The analyst should interactively vary the window width. Short windows reveal rapid membership changes but may fragment communities; longer windows reveal durable structures but can conceal splits or temporary collaborations.

For each snapshot, calculate:

• weighted co-authorship strength;
• publication count;
• weighted degree and within-community degree;
• betweenness or brokerage measures;
• number of recurring collaborators;
• venue composition;
• citation impact, where available.

---

3. Discover candidate overlapping communities

Use an overlapping community model, not a hard partition. Suitable approaches include link-community detection, mixed-membership stochastic block models, or overlapping modularity methods. The important design choice is that each author receives a membership weight for one or more communities in each time window.

Run the detection at several resolutions and parameter settings. Do not treat one algorithmic result as definitive. Instead, compare candidate solutions according to:

• number and size distribution of communities;
• internal collaboration density;
• stability under bootstrap resampling of papers or edges;
• persistence of communities across adjacent time windows;
• interpretability of their topical profiles;
• plausibility according to visualization-domain knowledge.

The analyst should use a community-resolution control to merge overly fragmented clusters or examine subcommunities within large ones. This follows the human-guided refinement logic of SOMFlow: Guided Exploratory Cluster Analysis.

Report community number in two forms:

• the number of active communities per time window;
• the number of distinct community lineages across the full 1990–2024 period.

This distinction prevents temporary splits or renamed continuations from being counted as entirely new communities.

---

4. Identify cores, leaders, and peripheral members

For every community and time window, classify members by both structural role and temporal persistence.

A practical classification is:

• Stable core: authors with sustained, high membership in the community over multiple adjacent windows and repeated collaborations with other core members.
• Leaders: stable-core authors with high within-community collaboration strength, high publication contribution, or strong brokerage between the community and others.
• Peripheral contributors: authors with limited, short-lived, or low-strength membership.
• Boundary spanners: authors with meaningful membership in multiple communities or high cross-community collaboration.

Do not define leaders solely by publication count or citation count. Present a composite profile including:

• years active in the community;
• within-community weighted degree;
• number of repeated co-authors;
• cross-community brokerage;
• paper and citation impact;
• topical contribution.

The analyst should be able to inspect the underlying paper list for any proposed leader, especially when a high centrality score is driven by one large team paper or a short-lived collaboration.

---

5. Track community evolution and lineage events

Match communities between consecutive temporal snapshots using a combination of:

• weighted author-membership overlap;
• overlap of stable-core members;
• topical similarity;
• collaboration-structure similarity.

Use core overlap more heavily than peripheral overlap when deciding whether a community persists. A community whose peripheral membership changes substantially but whose core remains stable should normally be treated as a continuation.

Classify temporal events using explicit rules, subject to analyst review:

Event	Operational indication
Continuation	One community in period (t) strongly matches one in (t+1)
Growth / contraction	Continuation with substantial size change
Merge	Multiple prior communities substantially contribute to one later community
Split	One prior community substantially contributes to multiple later communities
Birth	A later community has no sufficient predecessor
Dissolution	A community has no sufficient successor
Temporary disappearance	No successor for a short gap, followed by a strong later match
Reorientation	Stable core persists but topical profile changes substantially

Visualize the result as a community lineage diagram: communities appear as bands over time, band width encodes active membership or publication volume, and links encode continuation, splitting, and merging. Membership overlap should be visible separately from lineage; otherwise, a few mobile authors can falsely imply a merger.

Use the dynamic-network perspective of Reducing Snapshots to Points as a complementary overview. Represent each temporal snapshot by a vector of network and community features—such as community-size distribution, core persistence, overlap level, modularity, and topic mixture—and project snapshots into a “network-state map.” This can reveal periods with unusually rapid restructuring, such as the emergence of new venues, interdisciplinary expansion, or changes in collaboration scale.

---

6. Characterize each community’s research foci

For every community–time-window pair, build a weighted document corpus from papers authored by its members.

When an author belongs to multiple communities, fractionally allocate the paper to each community according to that author’s membership weight. Retain a second view of full paper membership so the analyst can inspect interdisciplinary papers without forcing them into one topic.

Apply an initial topic model, preferably NMF or another interpretable topic method, to titles, abstracts, and keywords. Then use the interactive refinement strategy from Human-Steered Topic Modelling:

• merge redundant topics;
• split topics that mix distinct research areas;
• remove generic terms;
• seed missing domain concepts using key terms or selected papers;
• compare topic assignments against paper-level evidence.

For each community, show:

• its top topics and representative keywords;
• representative papers for each topic;
• topic prevalence over time;
• emerging, declining, and persistent topics;
• topic diversity and topical concentration;
• topics shared with other communities.

If citation links are available, add citation-flow summaries: which communities cite one another, which external fields they draw from, and whether a topical change coincides with new citation connections.

---

7. Provide coordinated visual analysis views

A research analyst should work across linked views rather than rely on a single network picture.

View	Main purpose
Community lineage view	Identify persistence, births, splits, merges, and dissolutions
Temporal membership matrix / author ribbons	Inspect individual affiliation histories and multi-community participation
Overlap view	Compare communities through shared members; show pairwise overlap and its temporal change
NodeTrix time-slice view	Inspect selected communities as matrices while retaining inter-community links, following NodeTrix
Core–periphery view	Compare leaders, stable cores, peripheral members, and brokers
Topic timeline / heatmap	Examine each community’s topical profile and topic drift
Paper and citation inspector	Validate patterns through titles, abstracts, keywords, authorship teams, and citations
Snapshot-state projection	Identify stable eras, recurrent structures, and anomalous periods, following Reducing Snapshots to Points
Provenance panel	Record community parameters, topic edits, event interpretations, and analyst annotations, following SOMFlow

For overlap analysis, the analyst should be able to select two or more communities and see:

• shared authors;
• authors with strong membership in both;
• shared papers;
• common topics;
• cross-community collaboration edges;
• whether overlap is stable, increasing, or limited to a particular period.

This follows the set-comparison reasoning used in StratomeX, adapted from patient-set overlap to author-community overlap.

---

8. Investigate major evolutionary events with evidence

When the lineage view identifies a split, merge, reorientation, or dissolution, open an event-focused comparison workspace:

1. Compare the predecessor and successor communities’ stable cores.
2. Identify members who stayed, left, joined, or became boundary spanners.
3. Compare their collaboration matrices before and after the event.
4. Compare topic distributions, representative papers, and key terms.
5. Inspect venue shifts and, if available, citation flows.
6. Review a timeline of highly central papers and new collaborations around the event.
7. Record an analyst interpretation of the likely cause.

Potential interpretations should be framed as evidence-backed hypotheses, not automatic causal claims. For example:

• a merge may coincide with increased collaboration between previously separate methodological and application-oriented groups;
• a split may reflect divergence into visual analytics versus domain-specific visualization work;
• a topical reorientation may coincide with new data types, algorithms, venues, or influential papers;
• a dissolution may be an artifact of sparse publication periods rather than the end of an intellectual community.

---

9. Iterate, validate, and document uncertainty

The analyst should repeat detection and tracking under reasonable alternatives:

• different rolling-window widths;
• different edge-weighting schemes;
• different community resolutions;
• inclusion versus exclusion of weak, one-off ties;
• alternative topic-model sizes and preprocessing choices.

Flag conclusions that are sensitive to these choices. In particular, distinguish:

• robust communities, visible across parameter settings;
• tentative communities, present only at a narrow resolution or in sparse periods;
• uncertain lineage events, where several predecessors or successors are similarly plausible.

Maintain an analysis history containing parameters, manually adjusted community boundaries, topic edits, event labels, and supporting paper-level evidence. This makes the resulting intellectual-history account reproducible and open to later revision.

---

Expected analytical outcomes

The workflow should yield:

1. A defensible estimate of the number of active communities and long-term community lineages.
2. For each community, a profile of its stable core, leaders, peripheral members, and boundary spanners.
3. A temporally resolved account of membership overlap and researcher mobility.
4. Topic profiles showing each community’s major research foci and topical evolution.
5. A reviewed event history of community births, merges, splits, reorientations, and dissolutions.
6. Evidence-backed interpretations linking structural changes to papers, collaborations, topics, venues, and citation patterns.

Listing relevant approaches

I need to compile a list of relevant papers from the response, and it’s important to be honest and concise. The response mentioned NodeTrix, Reducing Snapshots, Human-Steered, SOMFlow, and StratomeX. I should focus on those five, as the workflow explicitly combines ideas from them. Additionally, I need to ensure that I cite these within the text appropriately, sticking to the request for English and highlighting their relevance.

Relevant library papers and the approaches used:

1. Reducing Snapshots to Points: A Visual Analytics Approach to Dynamic Network Exploration
   Stef van den Elzen et al., 2016
   - Used for representing time-windowed co-authorship networks as evolving network states.
   - Informs the snapshot-state projection for finding stable periods, transitions, and unusual restructuring.

2. NodeTrix: A Hybrid Visualization of Social Networks
   Nathalie Henry et al., 2007
   - Used for detailed inspection of dense collaboration communities while preserving links between communities.
   - Informs the proposed matrix-plus-node-link view for core structure, internal collaboration, and boundary-spanning authors.

3. Human-Steered Topic Modelling / UTOPIAN: User-Driven Topic Modeling Based on Interactive Nonnegative Matrix Factorization
   J. Choo et al., 2013
   - Used for characterizing communities’ research foci from titles, abstracts, and keywords.
   - Informs analyst-guided topic refinement: merging, splitting, seeding, and editing topics rather than accepting an automatic topic model uncritically.

4. SOMFlow: Guided Exploratory Cluster Analysis with Self-Organizing Maps and Analytic Provenance
   D. Sacha et al., 2018
   - Used for the iterative, human-guided approach to community detection and resolution selection.
   - Informs maintaining provenance of parameter choices, community refinements, and interpretations across analysis iterations.

5. Guided Visual Exploration of Genomic Stratifications in Cancer (StratomeX)
   M. Streit et al., 2014
   - Used for comparing overlapping groups and tracing membership across alternative or evolving groupings.
   - Informs the community-overlap views, membership comparisons, and linked inspection of shared members and associated characteristics.

6. Spatio-temporal Analysis and Modelling
   N. Andrienko and G. Andrienko, 2013
   - Relevant to iterative clustering of temporal profiles and analyst validation of cluster coherence.
   - Informs the recommendation to inspect and refine temporal community groupings rather than treat clustering as final.

7. Visual Analysis of Multivariate State Transition Graphs
   A. J. Pretorius and J. J. van Wijk, 2006
   - Relevant to analyzing transitions among abstract states.
   - Informs the explicit treatment of community continuations, splits, merges, births, and dissolutions as interpretable evolution events.

8. Jigsaw: Supporting Investigative Analysis through Interactive Visualization
   J. Stasko et al., 2008
   - Relevant to grounding high-level patterns in original textual evidence.
   - Informs the paper-level inspector, where analysts validate community-topic interpretations using underlying titles, abstracts, authorship records, and citations.

Evaluating model outputs

I'm considering the "build model" output and whether it can generate multiple typical models, particularly concerning community units and their features. First, I should prepare by defining a clean paper corpus from raw sources, ensuring that I can filter or normalize them properly. I also think about how to include author IDs for standardization and how to organize community members according to topic clustering. It appears that I need a main loop for build model to refine and assess topics dynamically. This approach supports community solutions and interpretable topics.

Developing community models

I’m looking at how to build a community model that uses F network and S, alongside D community instances. I think it’ll be important to define lineage over time after acceptance. The topology and temporal settings need specific analyst input, so I should consider separate predefined windows to draw from paper outputs. To define authors, I could group them from the papers to create a D_authors set. I might also think about producing canonical authors and individual network snapshots for better time embedding.

Defining time context

I’m trying to define D_time_context based on elementary time, but there’s more to consider. I might need to contextualize T arrangement in terms of lineage. The community lineage seems linked but isn’t finalized—maybe F can be a candidate? It feels important to arrange community instances within time lanes, matching them based on membership. Ultimately, I need to produce A_lineage, which will help visualize everything! Each piece matters in clarifying the overall context.

Considering model synthesis

I'm focusing on the accepted model and suggesting a final synthesis involving human interpretation. It could be useful to create an evidence exploration loop that determines event candidates and their defining units, as they have real event patterns. I think visualizing detailed event evidence allows for better assessment. However, it feels a bit unclear as I try to distinguish potential explanations from causal ones. I could benefit from an investigative loop on selected event sets.

Considering calculations and assessments

I need to focus on the input S and output K for assessing candidates and updating S. The description should mention that a machine calculates robustness while an analyst judges the results. I'm thinking about the hybrid role of the actor in this process. Outside the loop, I recognize the need for the community model, possibly called M_community_model, to reflect the latest updates. It feels like outlining all components for better clarity!

Considering citation sources

I’m pondering the citation requirements, specifically if they need to include links. It seems like there shouldn't be any issues, but it's better to confirm. I need to ensure that I gather relevant sources with appropriate titles, abstracts, and keywords. This way, I can provide helpful information while addressing whatever needs the user has regarding citations. It’s important to stay organized and thorough in my approach!

Finalizing assignment adjustments

I'm assessing the current configuration for the assignment. It looks like the result indicates that "refinement is required," but I wonder if maybe I'm not actually updating the output effectively. I might just finalize it without further revisions if that’s the case. It's a bit tricky to figure out the right balance here! But let's make sure everything's clear and ready to go.

workflow dynamic-vis-community-analysis
  template: loop(define-unit -> assess) ->
            loop(define-unit (time-windowing) -> define-unit (network construction) ->
                 characterise -> build-model -> define-unit (overlapping communities) ->
                 characterise -> define-unit (community corpora) -> build-model ->
                 characterise -> contextualise -> visualise -> abstract -> assess) ->
            define-unit (community lineages) -> characterise -> abstract ->
            visualise ->
            loop(define-unit (event selection) -> visualise ->
                 generate-knowledge -> assess) ->
            generate-knowledge
  description: "Human-in-the-loop workflow for identifying, validating, and
    interpreting evolving overlapping research communities in IEEE VIS
    co-authorship data, including stable cores, leaders, topical foci,
    membership overlap, and lineage events."

# ============================================================================
# INPUTS AND INITIAL ANALYTIC CONTROL
# ============================================================================

artifact D_papers_raw : entities
  origin: given
  internal structure: elementary
  embedment: time
  features:
    - id: f_paper_id
      value structure: atomic
      value type: reference
      description: "Unique paper record identifier"
    - id: f_year
      value structure: atomic
      value type: temporal
      description: "Publication year, 1990--2024"
    - id: f_venue
      value structure: atomic
      value type: categorical
      description: "VIS, InfoVis, VAST, or SciVis venue"
    - id: f_authors
      value structure: list
      value type: reference
      description: "DBLP-disambiguated author identifiers, retaining raw names"
    - id: f_title_abstract_keywords
      value structure: vector
      value type: text
      description: "Paper title, abstract, and author keywords"
    - id: f_citation_count
      value structure: atomic
      value type: numeric
      description: "Available citation count for the paper"
  description: "IEEE VIS-family paper records containing authorship, bibliographic,
    textual, and citation information."

artifact K_research_questions : knowledge(D_papers_raw)
  origin: given
  representation form: "analytic questions and interpretation criteria"
  description: "Questions concerning the number, stability, leadership,
    topical foci, overlap, and merge/split/dissolution history of
    collaboration communities."

artifact S_curation : specification
  origin: given
  representation form: "data-quality rules"
  description: "Rules for duplicate-paper handling, DBLP identifier checks,
    author-list validation, missing-text treatment, and retention of raw
    author strings for inspection."

artifact S_analysis_design_initial : specification
  origin: given
  representation form: "configuration schema"
  description: "Initial choices for rolling temporal-window width and stride,
    fractional co-authorship weighting, weak-tie treatment, overlapping
    community method and resolution range, temporal smoothing, robustness
    tests, topic-model configuration, and event-matching thresholds."

artifact S_event_review_initial : specification
  origin: given
  representation form: "review priorities"
  description: "Criteria for selecting lineage events for detailed review,
    prioritising high-impact, structurally uncertain, or topically important
    merges, splits, reorientations, and dissolutions."

assign:
  D_papers_current := D_papers_raw
  S_analysis_design := S_analysis_design_initial
  S_event_review := S_event_review_initial

# ============================================================================
# DATA CURATION LOOP
# ============================================================================

loop L_data_curation:
  purpose: "Create an auditable paper corpus with reliable author identities
    and usable textual metadata."
  until: "Paper records, author identities, and retained text are adequate
    for longitudinal collaboration and topic analysis."
  body:

    transform T_curate_papers :
      intent: define-unit
      manner: "validate, deduplicate, and normalise publication records"
      input: D_papers_current, S_curation
      output: D_papers_curated
      actor: hybrid
      description: "Resolve duplicate or incomplete records; validate DBLP
        author identifiers against raw names; retain provenance for exclusions;
        and normalise textual fields without removing meaningful domain terms."

    artifact D_papers_curated : entities
      internal structure: elementary
      embedment: time
      features:
        - id: f_author_ids
          value structure: list
          value type: reference
          description: "Validated DBLP author identifiers"
        - id: f_text
          value structure: vector
          value type: text
          description: "Cleaned title, abstract, and keyword text"
        - id: f_venue
          value structure: atomic
          value type: categorical
          description: "Publication venue"
        - id: f_citation_count
          value structure: atomic
          value type: numeric
          description: "Citation count retained for impact summaries"
      description: "Curated VIS-family papers suitable for collaboration,
        topical, and temporal analysis."

    transform T_assess_curation :
      intent: assess
      manner: "audit data completeness and identity reliability"
      input: D_papers_curated, S_curation, K_research_questions
      output: K_curation_quality
      actor: hybrid
      description: "Inspect missing metadata, anomalous author teams,
        suspicious name--identifier mappings, and venue/year coverage."

    artifact K_curation_quality : knowledge(D_papers_curated)
      representation form: "quality judgment"
      description: "Assessment of whether the curated corpus is sufficiently
        complete, identity-consistent, and temporally representative."

    assign:
      D_papers_current := D_papers_curated

    if K_curation_quality indicates unresolved data problems:
      then:
        continue
      else:
        exit loop L_data_curation

end loop L_data_curation

# ============================================================================
# COMMUNITY DISCOVERY, TOPIC CHARACTERISATION, AND DESIGN-REFINEMENT LOOP
# ============================================================================

loop L_community_design:
  purpose: "Develop a robust and interpretable representation of evolving,
    overlapping collaboration communities and their research foci."
  until: "Community structure is robust across reasonable alternatives,
    temporally coherent, topically interpretable, and adequate for the
    research questions."
  body:

    transform T_define_time_windows :
      intent: define-unit
      manner: "overlapping rolling time-window partitioning"
      input: D_papers_current, S_analysis_design
      output: D_time_windows
      actor: hybrid
      description: "Create annual and rolling analysis windows according to
        the current design, allowing the analyst to balance temporal detail
        against community stability."

    artifact D_time_windows : entities
      internal structure: elementary
      embedment: time
      features:
        - id: f_window_extent
          value structure: atomic
          value type: temporal
          description: "Start year, end year, and duration of each window"
      description: "Temporally ordered publication windows used to construct
        dynamic collaboration networks."

    transform T_construct_collaboration_networks :
      intent: define-unit
      manner: "author aggregation and time-windowed co-authorship construction"
      input: D_papers_current, D_time_windows, S_analysis_design
      output: D_author_windows, D_collaboration_networks
      actor: machine
      description: "Create author-by-window analysis units and one weighted
        co-authorship network per time window; retain links to contributing
        papers and apply fractional contribution rules for multi-author papers."

    artifact D_author_windows : entities
      internal structure: elementary
      embedment: {set, time}
      features:
        - id: f_author_id
          value structure: atomic
          value type: reference
          description: "DBLP author identifier"
        - id: f_window_id
          value structure: atomic
          value type: reference
          description: "Temporal window identifier"
        - id: f_contributing_papers
          value structure: list
          value type: reference
          description: "Papers written by the author in this window"
      description: "Author presences indexed by researcher and analysis window."

    artifact D_collaboration_networks : entities
      internal structure: formation
      embedment: time
      features:
        - id: f_window_id
          value structure: atomic
          value type: reference
          description: "Temporal window represented by the network snapshot"
      description: "Time-indexed weighted co-authorship network snapshots."

    transform T_characterise_collaboration :
      intent: characterise
      manner: "compute weighted network and author-activity descriptors"
      input: D_collaboration_networks, D_author_windows, D_papers_current,
             S_analysis_design
      output: F_collaboration_relations, F_author_activity
      actor: machine
      description: "Compute fractional co-authorship weights, repeated-tie
        statistics, publication and citation summaries, local connectivity,
        and brokerage descriptors."

    artifact F_collaboration_relations : feature(D_collaboration_networks)
      value structure: relational configuration
      value type: numeric
      representation form: "weighted undirected co-authorship graphs"
      description: "Collaboration edges and weights for every temporal network."

    artifact F_author_activity : feature(D_author_windows)
      value structure: vector
      value type: numeric
      description: "Publication volume, citation summary, collaboration strength,
        repeated-collaborator count, and network-centrality descriptors."

    transform T_fit_community_model :
      intent: build-model
      manner: "fit dynamic overlapping community model"
      input: D_collaboration_networks, F_collaboration_relations,
             F_author_activity, S_analysis_design
      output: M_community_model
      actor: machine
      description: "Fit an overlapping, temporally regularised community model
        at the selected resolution; estimate probabilistic or weighted
        membership rather than forcing each author into one community."

    artifact M_community_model : model(D_collaboration_networks,
        F_collaboration_relations)
      model type: "dynamic overlapping community model"
      representation form: "time-indexed membership-weight matrices and
        community-interaction parameters"
      description: "Model of evolving co-authorship communities with
        multi-community author affiliation."

    transform T_define_candidate_communities :
      intent: define-unit
      manner: "extract overlapping communities from membership model"
      input: D_author_windows, M_community_model, S_analysis_design
      output: D_candidate_communities, F_author_membership
      actor: machine
      description: "Create candidate community instances for each time window
        and assign each author-window a membership weight in zero, one, or
        multiple communities."

    artifact D_candidate_communities : entities
      internal structure: group/cluster
      embedment: time
      features:
        - id: f_window_id
          value structure: atomic
          value type: reference
          description: "Window in which the community instance is active"
        - id: f_members
          value structure: list
          value type: reference
          description: "Author-window members with non-zero affiliation"
      description: "Candidate overlapping collaboration communities in each
        temporal window."

    artifact F_author_membership : feature(D_author_windows)
      value structure: matrix
      value type: numeric
      representation form: "author-window by community membership-weight matrix"
      description: "Weighted affiliations supporting simultaneous membership
        in multiple communities."

    transform T_characterise_candidate_evolution :
      intent: characterise
      manner: "match adjacent communities by core, membership, and topical proxy"
      input: D_candidate_communities, F_author_membership,
             F_collaboration_relations, S_analysis_design
      output: F_candidate_lineage
      actor: machine
      description: "Estimate candidate continuation links between communities
        in adjacent windows, giving greater weight to persistent high-membership
        authors than to transient peripheral members."

    artifact F_candidate_lineage : feature(D_candidate_communities)
      value structure: relational configuration
      value type: numeric
      representation form: "directed weighted links between adjacent time windows"
      description: "Candidate predecessor--successor relations among community
        instances."

    transform T_define_community_corpora :
      intent: define-unit
      manner: "fractional paper allocation by community membership"
      input: D_papers_current, D_candidate_communities, F_author_membership,
             S_analysis_design
      output: D_community_corpora
      actor: machine
      description: "Create a document corpus for each community-window pair.
        Papers with authors spanning multiple communities are fractionally
        represented while retaining links to their complete author teams."

    artifact D_community_corpora : entities
      internal structure: group/cluster
      embedment: time
      features:
        - id: f_papers
          value structure: list
          value type: reference
          description: "Papers contributing to the community corpus"
        - id: f_fractional_contribution
          value structure: vector
          value type: numeric
          description: "Community-specific paper contribution weights"
      description: "Time-indexed weighted paper corpora associated with
        candidate communities."

    transform T_fit_topic_model :
      intent: build-model
      manner: "fit interpretable topic model"
      input: D_community_corpora, S_analysis_design
      output: M_topic_model
      actor: hybrid
      description: "Fit a topic model to titles, abstracts, and keywords using
        analyst-controlled preprocessing, topic granularity, and domain-term
        constraints."

    artifact M_topic_model : model(D_community_corpora)
      model type: "topic model"
      representation form: "interpretable document-topic and topic-term matrices"
      description: "Model of research foci expressed in community-associated
        paper corpora."

    transform T_characterise_research_foci :
      intent: characterise
      manner: "infer topic mixtures and representative papers"
      input: D_candidate_communities, D_community_corpora, M_topic_model
      output: F_topic_focus, F_representative_papers
      actor: machine
      description: "Associate each community with topic prevalence, salient
        terms, temporal topic trajectories, and representative papers."

    artifact F_topic_focus : feature(D_candidate_communities)
      value structure: matrix
      value type: numeric
      representation form: "community-window by topic prevalence matrix"
      description: "Research-focus mixture for each candidate community over time."

    artifact F_representative_papers : feature(D_candidate_communities)
      value structure: list
      value type: reference
      description: "High-contribution papers supporting each community topic
        interpretation."

    transform T_contextualise_candidate_lineages :
      intent: contextualise
      manner: "time-aligned lineage layout"
      input: D_candidate_communities, F_candidate_lineage, F_topic_focus
      output: D_lineage_canvas, A_candidate_lineages
      actor: machine
      description: "Create a time-aligned layout in which community instances
        occupy temporal lanes and candidate lineage links are visually traceable."

    artifact D_lineage_canvas : entities
      internal structure: elementary
      features:
        - id: f_layout_dimensions
          value structure: atomic
          value type: numeric
          description: "Two layout dimensions: time and community lane"
      description: "Computed reference space for community-lineage display."

    artifact A_candidate_lineages : arrangement(D_candidate_communities)
      context: D_lineage_canvas
      principle: "time-aligned placement with lineage-aware lane ordering"
      description: "Arrangement of community instances for inspecting
        continuity, branching, merging, and topical change."

    transform T_visualise_candidates :
      intent: visualise
      manner: "coordinated community-evolution dashboard"
      input: A_candidate_lineages, D_collaboration_networks,
             D_candidate_communities, F_author_membership,
             F_candidate_lineage, F_topic_focus, F_representative_papers,
             F_author_activity
      output: V_candidate_dashboard
      actor: machine
      description: "Provide linked lineage, overlap, network, author-history,
        topic-timeline, and paper-inspection views for interactive analysis."

    artifact V_candidate_dashboard : visualisation(A_candidate_lineages,
        D_collaboration_networks, F_author_membership, F_topic_focus)
      layout: "coordinated lineage timeline, overlap matrix, NodeTrix-style
        selected-network view, author membership ribbons, and topic heatmap"
      form: "bands, links, matrices, node-link marks, timelines, and text lists"
      encoding: "band width shows community size or publication volume; links
        show lineage strength; membership weight controls opacity; topic
        prevalence controls colour; selected authors and papers are linked
        across all views"
      description: "Interactive evidence display for judging community
        boundaries, overlap, temporal persistence, and research foci."

    transform T_abstract_candidate_patterns :
      intent: abstract
      manner: "perception-and-interpretation of community structure"
      input: V_candidate_dashboard, D_candidate_communities,
             F_author_membership, F_candidate_lineage, F_topic_focus
      output: P_candidate_structure
      actor: human
      description: "Identify tentative stable cores, boundary spanners,
        suspicious fragmentation, potential lineage events, and meaningful
        topic labels from the coordinated views."

    artifact P_candidate_structure : pattern(D_candidate_communities,
        F_author_membership, F_topic_focus)
      representation form: "annotated candidate communities and topical labels"
      description: "Analyst interpretations of candidate community structure,
        overlap, and research-focus regularities."

    transform T_assess_candidate_design :
      intent: assess
      manner: "evaluate robustness, temporal coherence, and interpretability"
      input: V_candidate_dashboard, M_community_model, M_topic_model,
             P_candidate_structure, K_research_questions, S_analysis_design
      output: K_candidate_quality, S_analysis_design_r
      actor: hybrid
      description: "Compare parameter and resampling alternatives; assess
        community cohesion, stability, overlap plausibility, topic coherence,
        and whether apparent events are supported by underlying papers."

    artifact K_candidate_quality : knowledge(D_candidate_communities,
        M_community_model, M_topic_model)
      representation form: "quality judgment"
      description: "Assessment of whether candidate communities and topics are
        robust, interpretable, and sufficient for substantive analysis."

    artifact S_analysis_design_r : specification
      representation form: "revised configuration schema"
      description: "Updated choices for temporal aggregation, edge weighting,
        community resolution, smoothing, topic granularity, and lineage
        thresholds based on the assessment."

    if K_candidate_quality indicates refinement needed:
      then:
        assign:
          S_analysis_design := S_analysis_design_r
      else:
        exit loop L_community_design

end loop L_community_design

# ============================================================================
# ESTABLISH FINAL COMMUNITY LINEAGES, ROLES, AND EVOLUTION EVENTS
# ============================================================================

transform T_define_community_lineages :
  intent: define-unit
  manner: "assemble temporally linked community instances into lineages"
  input: D_candidate_communities, F_candidate_lineage, K_candidate_quality,
         S_analysis_design
  output: D_community_lineages, F_lineage_membership
  actor: hybrid
  description: "Confirm continuation relations and form distinct community
    lineages across 1990--2024, preserving uncertainty where multiple
    predecessors or successors remain plausible."

artifact D_community_lineages : entities
  internal structure: sequence
  embedment: set
  features:
    - id: f_lineage_instances
      value structure: list
      value type: reference
      description: "Ordered community instances belonging to the lineage"
    - id: f_lifetime
      value structure: atomic
      value type: temporal
      description: "First and last active analysis windows"
  description: "Longitudinal community identities composed of ordered
    community instances."

artifact F_lineage_membership : feature(D_candidate_communities)
  value structure: atomic
  value type: reference
  description: "Lineage identifier assigned to each community instance."

transform T_characterise_roles_and_overlap :
  intent: characterise
  manner: "compute persistence, leadership, brokerage, and overlap profiles"
  input: D_author_windows, D_community_lineages, F_author_membership,
         F_author_activity, F_collaboration_relations, F_lineage_membership
  output: F_core_persistence, F_leadership, F_author_roles,
          F_community_overlap
  actor: machine
  description: "Identify stable-core membership from sustained high affiliation
    and repeated internal ties; score leadership from within-community
    contribution and brokerage; quantify overlap and author mobility."

artifact F_core_persistence : feature(D_author_windows)
  value structure: atomic
  value type: numeric
  description: "Persistence score for stable-core participation in a lineage."

artifact F_leadership : feature(D_author_windows)
  value structure: atomic
  value type: numeric
  description: "Leadership score combining internal collaboration strength,
    sustained contribution, publication activity, and brokerage."

artifact F_author_roles : feature(D_author_windows)
  value structure: list
  value type: categorical
  description: "Roles such as stable core, leader, peripheral contributor,
    and boundary spanner; multiple roles may apply across communities."

artifact F_community_overlap : feature(D_candidate_communities)
  value structure: matrix
  value type: numeric
  representation form: "time-indexed pairwise overlap matrix"
  description: "Shared-member and weighted-membership overlap between
    communities in each time window."

transform T_abstract_evolution_events :
  intent: abstract
  manner: "classify community lineage events"
  input: D_community_lineages, F_candidate_lineage, F_lineage_membership,
         F_author_membership, F_core_persistence, F_topic_focus
  output: P_evolution_events
  actor: hybrid
  description: "Identify and label continuations, growth, contraction, births,
    dissolutions, temporary disappearances, mergers, splits, and topical
    reorientations; distinguish candidate events from robust events."

artifact P_evolution_events : pattern(D_community_lineages,
    F_candidate_lineage, F_topic_focus)
  representation form: "time-stamped event records with confidence annotations"
  description: "Structural and topical evolution events in the VIS research
    community."

transform T_visualise_final_structure :
  intent: visualise
  manner: "interactive community-history dashboard"
  input: D_community_lineages, D_candidate_communities,
         F_author_membership, F_core_persistence, F_leadership,
         F_author_roles, F_community_overlap, F_topic_focus,
         F_representative_papers, P_evolution_events
  output: V_community_history
  actor: machine
  description: "Present final community lineages and enable drill-down from
    high-level evolution patterns to authors, collaborations, papers, and
    topical evidence."

artifact V_community_history : visualisation(D_community_lineages,
    F_author_roles, F_community_overlap, F_topic_focus, P_evolution_events)
  layout: "linked community lineage view, role-ranked author table, overlap
    matrix, membership ribbons, topic evolution heatmap, and event detail panel"
  form: "bands, flow links, matrices, ranked lists, timelines, and paper cards"
  encoding: "lineage bands encode active membership or publication volume;
    core and leader scores control author ranking; event markers encode type
    and confidence; topics encode colour and relative area"
  description: "Final visual workspace for answering community-structure and
    evolution questions with traceable evidence."

# ============================================================================
# EVENT-FOCUSED INTERPRETATION LOOP
# ============================================================================

artifact K_event_log_initial : knowledge(P_evolution_events)
  origin: given
  representation form: "annotated evidence log"
  description: "Initially empty or externally supplied record of prior
    interpretations, uncertainties, and domain context for evolution events."

assign:
  K_event_log := K_event_log_initial

loop L_event_interpretation:
  purpose: "Develop evidence-backed explanations for consequential or
    uncertain community evolution events."
  until: "All salient events have been reviewed to a level appropriate for
    the research questions, and remaining uncertainty is explicitly recorded."
  body:

    transform T_select_events_for_review :
      intent: define-unit
      manner: "filter and prioritise event instances"
      input: P_evolution_events, S_event_review, K_event_log
      output: D_events_for_review
      actor: human
      description: "Select the next merge, split, dissolution, reorientation,
        or uncertain lineage event for investigation."

    artifact D_events_for_review : entities
      internal structure: elementary
      embedment: time
      features:
        - id: f_event_type
          value structure: atomic
          value type: categorical
          description: "Evolution-event category"
        - id: f_event_confidence
          value structure: atomic
          value type: numeric
          description: "Confidence in the event classification"
      description: "Selected event instances requiring detailed interpretation."

    transform T_visualise_event_evidence :
      intent: visualise
      manner: "before-and-after comparative evidence display"
      input: D_events_for_review, V_community_history,
             F_core_persistence, F_leadership, F_author_membership,
             F_topic_focus, F_representative_papers, F_author_activity
      output: V_event_evidence
      actor: machine
      description: "Compare predecessor and successor cores, entrants and
        leavers, collaboration structure, topic trajectories, representative
        papers, venue composition, and citation impact around each event."

    artifact V_event_evidence : visualisation(D_events_for_review,
        F_author_membership, F_topic_focus, F_representative_papers)
      layout: "aligned pre-event and post-event views with linked paper evidence"
      form: "community matrices, author flows, topic comparisons, and paper lists"
      encoding: "membership and core scores identify retained and changing
        members; topic proportions compare research foci; selected papers
        provide textual evidence"
      description: "Evidence workspace for interpreting why a structural
        community event may have occurred."

    transform T_formulate_event_interpretations :
      intent: generate-knowledge
      manner: "formulate qualified explanatory hypotheses"
      input: D_events_for_review, V_event_evidence, K_event_log,
             K_research_questions
      output: K_event_log_r
      actor: human
      description: "Record evidence-backed, non-causal interpretations such as
        topical divergence, new cross-group collaboration, researcher mobility,
        venue shifts, or sparse-data artifacts; retain alternative explanations."

    artifact K_event_log_r : knowledge(D_events_for_review, V_event_evidence)
      representation form: "annotated hypotheses with evidence and uncertainty"
      description: "Cumulative interpretations of reviewed evolution events,
        including supporting papers, structural evidence, and unresolved issues."

    transform T_assess_event_evidence :
      intent: assess
      manner: "judge explanatory sufficiency and prioritise further review"
      input: K_event_log_r, V_event_evidence, S_event_review
      output: K_event_evidence_quality, S_event_review_r
      actor: human
      description: "Assess whether each interpretation is sufficiently supported,
        whether alternative time windows or parameter settings should be checked,
        and which event should be reviewed next."

    artifact K_event_evidence_quality : knowledge(K_event_log_r)
      representation form: "quality judgment"
      description: "Judgment of evidential sufficiency, remaining uncertainty,
        and whether further event analysis is required."

    artifact S_event_review_r : specification
      representation form: "revised review priorities"
      description: "Updated event-selection priorities and evidence requirements."

    assign:
      K_event_log := K_event_log_r

    if K_event_evidence_quality indicates further event review needed:
      then:
        assign:
          S_event_review := S_event_review_r
      else:
        exit loop L_event_interpretation

end loop L_event_interpretation

# ============================================================================
# SYNTHESIS
# ============================================================================

transform T_synthesise_findings :
  intent: generate-knowledge
  manner: "formulate-statements and comparative community profiles"
  input: V_community_history, D_community_lineages, F_core_persistence,
         F_leadership, F_author_roles, F_community_overlap, F_topic_focus,
         P_evolution_events, K_event_log, K_candidate_quality
  output: K_community_findings
  actor: human
  description: "Produce a traceable account of the number of communities and
    lineages, stable cores and leaders, research foci and their evolution,
    membership overlap, and evidence-qualified merge, split, reorientation,
    and dissolution interpretations."

artifact K_community_findings : knowledge(D_community_lineages,
    P_evolution_events, F_topic_focus)
  representation form: "community profiles, longitudinal findings, and
    uncertainty-qualified conclusions"
  description: "Final analytical findings about the intellectual structure
    and evolution of the IEEE VIS research community."

Summarizing departures

I need to make sure my response is concise and in English. It seems there might be a question about whether the ATWL code was fabricated. I'll focus on summarizing departures by pointing out structural transformations like loops and artifacts, as well as specifications and iterative assessments. Importantly, there’s no need to involve math in this summary, so I'll keep it straightforward. I’m keen to get this right!

The ATWL design departs from the natural-language workflow mainly by making iteration, analyst control, and evidence provenance explicit.

• Data curation becomes a separate loop.
  The natural-language version began with preparation; ATWL treats data-quality validation as an iterative loop that must be assessed before community analysis proceeds.

• Community detection and topic modelling are coupled in one refinement loop.
  Rather than a linear sequence of “detect communities, then characterize topics,” ATWL repeatedly revises temporal windows, network weighting, community-model settings, and topic-model settings until the result is robust and interpretable.

• Specifications are first-class artifacts.
  ATWL introduces explicit, revisable specifications for:

• temporal-window construction;
• co-authorship weighting;
• community resolution and smoothing;
• topic-model granularity;
• lineage-matching thresholds;

• event-review priorities.
  This makes analyst decisions inspectable and reproducible.

• Overlapping membership is represented formally.
  The ATWL workflow defines an author–window-by-community membership matrix, rather than describing overlap only conceptually. This supports simultaneous affiliation, mobility, and boundary-spanner analysis.

• Community lineage construction is separated from initial detection.
  Candidate communities are first detected per time window. Only after validation are they assembled into longer-term lineages. This avoids prematurely treating an algorithmic cluster in one period as a persistent intellectual community.

• Roles are computed after lineage establishment.
  Stable cores, leaders, peripheral contributors, and boundary spanners are derived from persistence, internal collaboration, activity, and brokerage measures after communities have been longitudinally linked.

• Event interpretation is a dedicated loop.
  Merges, splits, dissolutions, and topical reorientations are not merely detected and reported. ATWL creates an event-review cycle in which the analyst selects important events, compares pre- and post-event evidence, formulates qualified hypotheses, and records uncertainty.

• Knowledge generation is explicitly distinct from pattern recognition.
  ATWL distinguishes:

• detected patterns, such as a possible split or high membership overlap;
• assessed quality, such as robustness or interpretability;

• generated knowledge, such as an evidence-backed interpretation of why a community changed.

• Visualization is embedded at decision points.
  Instead of treating visualizations as a final dashboard, ATWL places visual inspection before assessment and interpretation steps: candidate-community review, final-structure exploration, and event-focused evidence comparison.

• Uncertainty and provenance are formal outputs.
  The final findings incorporate robustness judgments, event confidence, alternative explanations, and an analyst-maintained evidence log, rather than presenting a single definitive community history.