Data Transparency
How Pasal.id collects, verifies, and presents Indonesian statutory texts.
Primary data comes from the official national legislation portal. Secondary sources include BPK JDIH (69 regulation types), the Constitutional Court JDIH (MK decisions), and JDIHN (regional regulations). Every document is linked to its original source URL.
PDFs are processed through a structural parser (LiteParse/Unstructured) to extract the article hierarchy. Results are verified by Gemini 3 Flash. Regressions are tested against a manually curated golden test suite. Admins can re-review suspicious results.
Users can submit corrections via the /koreksi/[nodeId] form on every article. Corrections enter an admin review queue. Once approved, content is updated directly with a full audit trail. Rejected corrections include a reason.
Scrapers run periodically for each regulation type. The freshness window is 168 hours (1 week) per type — types scraped within the last week are skipped. New regulations are discovered automatically through continuous discovery.
OCR errors may occur for low-quality scanned PDFs. Amendment chains in complex omnibus laws may be incomplete. Search uses keyword-based full-text search, not semantic search. All content is informational and does not constitute legal advice.
Full technical details are available on the Methodology page.
Methodology Page