Bibliometricnetwork
Who grows the field, along which axes?
Turns citation, co-authorship, keyword and temporal flows into a research map.
RUN9B1C-44E0CORPUSBPA & endocrine · 1,284METHODbiblio · LeidenSTATUSexampleTS2026-06-14 14:08 UTC
Three analysis types, one research memory
Lacuna shows the logic of the research universe, not just routine analyses.
The product is not a tool; it is an evidence architecture that can become a paper.
Thematicstrategic map
What does the literature say — and what does it not?
Does not just list themes; it surfaces conceptual proximity, repetition and zones of silence.
RUN9B1C-44E0CORPUSBPA & endocrine · 1,284METHODthematic · co-wordSTATUSexampleTS2026-06-14 14:08 UTC
Metaforest
How strong is the evidence, how defensible the decision?
Takes meta-analysis out of a calculator output and binds it to a decision trail: choice, assumption, heterogeneity and provenance.
RUN9B1C-44E0CORPUSBPA & endocrine · 14 studiesMETHODmeta · random-effects (DL)STATUSexampleTS2026-06-14 14:08 UTC
Each card is an illustrative example of an engine output, shown small: a tri-colour identity keyline, the output kind, the figure, and a specimen footer under it. The corpus and figures are a representative sample, not a live run. Engine colours (navy · jade · orange) live only in the chrome — every data mark stays in the colour-blind-safe palette.
Intellectual Structure · Social Structure · Co-word Analysis & Thematic Map · Topic Modeling (LDA / NMF) · Network Meta-Analysis · Dose–Response Meta-Analysis
Co-citation network · Leiden · 3 communities
The intellectual structure, as a living network
Co-citation links resolve the literature into research communities, with the bridge publications that connect them highlighted. Hover a node to focus its neighbourhood; drag, scroll or use the controls to explore.
Co-citation network · Leiden · 3 communities
Workflow
One question should remain in the user's mind: where should my study be built?
1
The data universe is built
Records from sources like Scopus, WoS and Scholar are cleaned, versioned and turned into an analysable research corpus.
Bisphenol-A & endocrine outcomes
Cohort and case-control studies on BPA exposure and reproductive risk.
RUN9B1C-44E0CORPUSBPA & endocrine · 1,284 recordsMETHODingest · RIS parser v1.2STATUSexampleTS2026-06-14 14:08 UTC
Start an analysisopens the guided workspace — data to defensible result
Illustrative walkthrough — the corpus, figures and provenance strip are a representative example of a real run, not live results.
Platform philosophy
Any analysis can be run; only a defensible analysis gets published.
Lacuna's premium feel does not come from ornament. Every screen carries an academic audit layer that answers the question: how did you reach this result?
LACUNAlcn-2026-0612-r41 · frame v2 · D-017
Higgins & Thompson (2002) — I² heterogeneity
Random-effects model selected because I² = 62% → the studies-differ assumption holds.
Representative seal — every Lacuna output carries one.
- 01
Evidence aesthetics
Colour and graphics are not decor; they carry methodological status. Bronze is the gap, deep navy is evidence, teal is the analytic-discovery signal.
- 02
Decision ledger
Model selection, filtering, exclusion and reporting choices remain as a record you can defend later.
- 03
Publication discipline
Each output is not a screenshot; it is a portable academic object carrying provenance, version, source and method context.
- 04
Ecosystem bond
ScholarGate is the learning layer, PaperMind the topic-finding layer, Lacuna the evidence-building layer. Not one product — a family that is an academic operating system.