Skip to main content

Public sources by knowledge area

Beyond YouTube and manual upload, the Researcher profile can search 26 public sources, organized into 9 knowledge areas — none require registration or an API key. Results are downloaded and prepared for indexing automatically, exactly like any other document in the Repository.

Mapped from "Bases de Dados Abertas — Consolidação Ampliada de Endpoints de API" (Brazil's Ministry of Health, NDTI/DECIT/SCTIE team, Jul/2026). Only sources tested live, with substantive text content, are included — not metadata-only feeds.

  1. In the extraction modal, choose Public source instead of YouTube
  2. Select the knowledge area and the specific source
  3. Type a topic or keywords
  4. Choose how many results to download (up to 50)

Sources by area

General and multidisciplinary search engines

Don't belong to a single domain area — index output from any field.

SourceWhat it covers
arXivPreprints in physics, mathematics, computer science, and related fields
OpenAlexOpen catalog of global scientific output (~322 million works)
CrossrefDOI metadata — 150M+ academic works from any field
DataCiteDOIs for datasets, software, and research output
DOAJDirectory of open-access journals and articles
ZenodoCERN/OpenAIRE repository — datasets, code, and DOI-tagged artifacts
Open LibraryBook metadata and descriptions (an Internet Archive project)
data.europa.euEuropean Union open data portal
data.gov.ukUK government open data portal

Technology, AI, and data science

SourceWhat it covers
GitHubOpen-source repositories — README, topics, and language
Stack OverflowTechnical Q&A from the developer community
Hacker NewsAsk HN/Show HN discussions — tech and startup community

Economics, finance, and social sciences

SourceWhat it covers
Central Bank of BrazilEconomic and monetary series, reports, and statistics

Law, regulation, legislation, and government

SourceWhat it covers
Chamber of Deputies (Brazil)Federal legislative bills — full text in PDF when available
Federal Senate — LegislationAlready-enacted federal laws — official summary of each statute

Health, biology, and genetics

SourceWhat it covers
PubMedMedical and biomedical literature (NCBI/Entrez) — full abstract
Europe PMCBiomedical literature with a semantic annotation layer
ClinicalTrials.govWorldwide clinical trial registry (NIH/NLM)
UniProtProtein sequences and biological function
openFDAFDA-approved drug labels (indications, usage, warnings)

Clinical search is also available via FHIR/ResearchStudy — upload a FHIR Bundle (.json) through the Repository and Tusab recognizes and structures it automatically. Scoped to research studies only: never patient data, even as test data.

Earth sciences, climate, and space

SourceWhat it covers
NASA EarthdataCatalog of satellite and sensor data collections

Physics, chemistry, and materials

SourceWhat it covers
CERN Open DataLHC collision datasets

Cultural heritage, history, and archives

SourceWhat it covers
Art Institute of ChicagoCollection of works with curatorial narrative descriptions
The Metropolitan Museum of ArtCollection of 470,000+ works — curatorial metadata (culture, period, medium)

Anthropology, linguistics, and structured knowledge

SourceWhat it covers
Wikipedia (PT)Full-text search on the Portuguese-language Wikipedia
WiktionaryFree multilingual dictionary — etymology, pronunciation, and definitions

Why arXiv is in "general," not just technology/physics

arXiv doesn't "belong" to a single domain area — it covers physics, mathematics, computer science, and more. It sits alongside the other multidisciplinary search engines (OpenAlex, DataCite, DOAJ, Zenodo) instead of being buried inside a single area despite covering far more than that.

Extensibility

Each source is an independent module with the same contract (FONTE_META + buscar()) — adding a new source doesn't require touching the others. See tusab_engine/motor/fontes/ in the repository for the reference implementation.