# Universal Science Discovery Repository **The open-source knowledge engine for scientific unknowns and cross-domain discovery** [Bridges](cross-domain/) [Unknowns](unknowns-catalog/) [Hypotheses](hypotheses/) [Graph Nodes](docs/knowledge_graph.json) [Catalog & docs: CC BY 4.0](LICENSE) [Code & scripts: MIT](LICENSE-CODE) [CI](https://github.com/KR8ZYSHO3/Universal-Science-Discovery/actions) **[Live Dashboard](https://kr8zysho3.github.io/Universal-Science-Discovery/dashboard/) · [Knowledge Graph](https://kr8zysho3.github.io/Universal-Science-Discovery/dashboard/#knowledge-graph) · [Preprint](docs/preprint/usdr_preprint.md) · [Contribute**](CONTRIBUTING.md) --- ## The Problem Science has extraordinary infrastructure for recording what is *known* — 200M+ indexed papers, citation graphs, semantic search. It has almost no infrastructure for recording what is *unknown*. The frontier of human knowledge — the open problems, the cross-domain connections that haven't been made, the experiments that haven't been run — exists only in review article background sections, grant application paragraphs, and the informal knowledge of domain experts. It is not indexed. It is not version-controlled. It is not findable. This means: - A graduate student entering a field invests years to understand which problems are genuinely open - Two fields independently study the same mathematical phenomenon for decades without recognizing the connection - The most important questions in science are the hardest to find **USDR is built to fix this.** --- ## What We've Built A git-native, schema-validated, community-governed catalog of scientific unknowns, hypotheses, and cross-domain mathematical bridges — currently containing: | Metric | Count | Notes | | -------------------------- | --------- | --------------------------------------------------------------- | | **Cross-domain bridges** | **1,123** | Mathematical connections between fields that rarely communicate | | **Open unknowns** | **1,408** | Named, structured research gaps across 55+ disciplines | | **Falsifiable hypotheses** | **1,274** | Testable claims linked to specific unknowns | | **Knowledge graph nodes** | **3,857** | Interconnected across 4,517 edges | | **Pioneer profiles** | **18** | Scientists whose work seeded cross-domain bridges | | **Breakthrough gaps** | **24** | High-priority problems that would reshape entire fields | | **Orphan unknowns** | **0** | All unknowns connected to bridges or hypotheses | | **Schema errors** | **0** | All entries pass CI validation on every PR | Entries are YAML and schema-validated in CI; bridges and many records cite primary literature (DOIs/arXiv where applicable — see per-record `references`). The knowledge graph is rebuilt deterministically from source files on every push. --- ## What Makes a Bridge A USDR bridge is not a metaphor. It requires a **term-by-term mathematical mapping** between two fields — the same differential equation, order parameter, or information-theoretic quantity governing phenomena in both domains. **Example: Percolation Theory ↔ Ecological Fragmentation** | Statistical physics | Conservation biology | | -------------------------------------- | ------------------------------------------------------ | | Site percolation threshold p_c | Critical habitat coverage fraction (~59%) | | Giant connected component | Connected habitat network enabling population rescue | | Finite-size scaling correction | Threshold shift in smaller landscape patches | | Correlation length ξ (diverges at p_c) | Mean dispersal distance to maintain connectivity | | Universality class (2D, ν = 4/3) | Spatial connectivity independent of landscape geometry | The conservation biology literature has used percolation as a simulation tool since 1987. The exact analytical machinery — finite-size scaling, critical exponents, universality class proofs — has not crossed over. **That gap is what USDR makes explicit and actionable.** --- ## Discovery Engine USDR ships tooling that turns the knowledge graph into an active bridge-discovery system: ```bash # Find cross-domain papers from OpenAlex (250M+ papers) python scripts/harvesters/harvest_openalex.py --bridge-scan # Find the highest-priority domain pairs with no bridge yet python scripts/propose_bridges.py --top 15 # Find unknowns with no connections — prime contribution targets python scripts/find_orphan_unknowns.py # Audit catalog quality python scripts/audit_quality.py ``` The `harvest_openalex.py` script queries OpenAlex for papers that cite **two specific concept domains simultaneously** — the core primitive for automated bridge discovery at scale. PubMed and Semantic Scholar harvesters are also included. ### Wave Factory mode Wave Factory mode converts harvested candidates into schema-safe draft triples at scale: ```bash # Stage ranked bridge + unknown + hypothesis drafts python scripts/harvesters/wave_factory.py \ --top 30 \ --min-citations 50 \ --sources openalex,pubmed,semantic_scholar \ --output drafts/wave_factory # Validate staged records before promotion python scripts/harvesters/promote_wave_factory_batch.py --stage drafts/wave_factory ``` Drafts are intentionally staged (not auto-promoted) so human review remains the merge gate. --- ## Quick Start ```bash git clone https://github.com/KR8ZYSHO3/Universal-Science-Discovery.git cd Universal-Science-Discovery # View the live dashboard locally (path: dashboard/index.html — URL in dashboard/README.md) python -m http.server 8765 # Validate the entire catalog pip install pyyaml jsonschema python scripts/validate_schemas.py # Rebuild the knowledge graph pip install pyyaml networkx python scripts/build_graph.py ``` --- ## Catalog Structure ``` cross-domain/{domain-a}-{domain-b}/b-*.yaml ← 1123 bridges unknowns-catalog/{domain}/u-*.yaml ← 1408 unknowns hypotheses/active|validated|archived/h-*.yaml ← 1274 hypotheses pioneers/pioneer-*.yaml ← 18 pioneer profiles breakthrough-gaps/bg-*.yaml ← 24 breakthrough gaps phenomenology/p-*.yaml ← pre-formal observations schemas/ ← JSON Schema definitions scripts/ ← tooling and harvesters scripts/harvesters/ ← OpenAlex, PubMed, Semantic Scholar dashboard/ ← interactive D3 dashboard api/v1/ ← static JSON API ``` --- ## Contributing The simplest contribution is a new **unknown** — a named, structured research gap in your field. No code required, just YAML. ```yaml id: u-your-unknown-id title: "What mechanism allows X to Y without Z?" status: open summary: "A brief description of the gap and why it matters." disciplines: [your-field, adjacent-field] priority: high references: - doi: "10.XXXX/XXXXX" note: "Paper that defines the boundary of current knowledge" ``` See [CONTRIBUTING.md](CONTRIBUTING.md) for full instructions and [docs/QUICK_START_CONTRIBUTING.md](docs/QUICK_START_CONTRIBUTING.md) for a 30-minute first-contribution guide. **Open issues labeled `[good first issue](https://github.com/KR8ZYSHO3/Universal-Science-Discovery/issues?q=is%3Aopen+label%3A%22good+first+issue%22)`** include bridge stubs, domain unknowns, and documentation tasks across 8 scientific domains. --- ## Roadmap | Phase | Timeline | Goal | | -------------------------------------------- | --------- | ------------------------------------------------------------------------------------------ | | **Phase 0: Foundation** *(complete)* | 2026 | Governance · schemas · CI · seeded catalog · knowledge graph · contributor hub · automation | | **Phase 1: Discovery & adoption** *(active)* | 2026–2027 | arXiv DOI · public launch & outreach · first contributors · hackathon · custom domain | | **Phase 2: Momentum** | 2027–2028 | 10,000+ entries · institutional partners · AI co-pilot · 500 contributors | | **Phase 3: Acceleration** | 2029–2031 | 100,000+ entries · autonomous discovery agents · global research labs | | **Phase 4: Transformation** | 2032–2035 | The default interface through which researchers interact with the open frontier of science | --- ## Governance & Ethics - **Speculation is labeled.** `status: proposed` means plausible but unvalidated. `status: established` requires a published cross-disciplinary recognition. - **No fabricated citations.** All references must include a verifiable DOI or arXiv ID. - **Unknowns ≠ findings.** The system enforces terminological separation between what is known and what is not. - **Human review gates all merges.** CI catches schema errors; human domain expertise catches scientific errors. See [GOVERNANCE.md](GOVERNANCE.md), [ETHICS.md](ETHICS.md), and [docs/ETHICS_REPRODUCIBILITY_AND_DATA.md](docs/ETHICS_REPRODUCIBILITY_AND_DATA.md). --- ## Citation If you use USDR in research, please cite: > Shoemaker, B. and Contributors. *Universal Science Discovery Repository: Open Infrastructure for Tracking Scientific Unknowns and Cross-Domain Mathematical Bridges.* v1.1, May 2026. [https://github.com/KR8ZYSHO3/Universal-Science-Discovery](https://github.com/KR8ZYSHO3/Universal-Science-Discovery) A preprint is available at [docs/preprint/usdr_preprint.md](docs/preprint/usdr_preprint.md). --- ## License - **Catalog content** (YAML entries, documentation): [CC BY 4.0](LICENSE) - **Code** (scripts, dashboard, CI): [MIT](LICENSE) © 2026 Brandon Shoemaker and USDR Contributors