EthnoHack 2026 Traditional Knowledge Mining
Find proposed plant use records in published sources. Check the original passage, record your review, and export the evidence with its context intact.
No account needed. Use a real source from the starter corpus.
The example opens when your workspace connects.
Proposed extraction
Reported use Wounds
Hypericum perforatum L., Hypericaceae | Valsamochorto, Valsamo | Aerial parts | Immersed in olive oil | Oil topically applied | Wounds/cuts/burns (16)
Aerial parts
Immersed in olive oil
Table 3, selected rows · CC BY 4.0
DOI 10.3389/fphar.2023.1225580Published material.
Clearly bounded.
A starter corpus for inspection.
No independent accuracy validation claimed.
The workflow
One source trail, through every step.
Choose a starter paper or import permitted text. Citeflora saves the snapshot and citation, then proposes plant use records.
Text, supported files, or open literatureRead the highlighted passage. Inspect identity and uncertainty, correct fields when needed, and save your review decision.
Original extraction stays unchangedExplore reported associations and download your records with citations, review status, source spans, and recorded uncertainty.
PDF, DOCX or DOC, with citations and aligned fieldsContext is part of the finding
A plant name and a use are only the beginning. The original wording, preparation, citation, and uncertainty matter when someone checks the record later.
Citeflora keeps that context attached. Your corrections and decisions sit beside the extraction, with a review history you can inspect.
Build a collection you can check. Import a permitted passage, inspect the proposed records, and preserve the reasons behind your decisions.
Open a published source, inspect its passage, save a review, and export the result. The guided example uses the working application.
Before you begin
It proposes plant and reported use records from English text using a deterministic rule engine. Each candidate points to the exact passage in a saved source snapshot. You check the result, record a decision, and export the evidence. It is not a trained AI model.
Researchers, students, and data curators working with published ethnobotanical literature. It helps organize and inspect reported knowledge. It does not provide personal health advice or identify plants from photographs.
The starter corpus contains 43 selected table rows from 2 attributed CC BY studies, covering 18 source-reported taxa. It is a small working corpus, not the full papers, a comprehensive literature review, or new field research.
No. Acceptance records your review of the extraction against its source. A traditional use report does not establish clinical effectiveness or safety. Taxonomy suggestions also do not establish the identity of a historical specimen.
Yes. Paste permitted English text, upload PDF, DOC, DOCX, TXT, HTML or JSON, or use the six-column CSV template. You can also search Europe PMC and prepare a narrative import when its licence is supported. Add the citation and permission basis before saving.
Sources and review decisions are stored in Cloudflare D1 in a workspace associated with this browser. The session cookie lasts 30 days. There are no user accounts or cross-device sync. Download a formatted PDF, DOCX or DOC report to keep a portable copy.
Start with one source
Research records and reported associations. No treatment recommendations.