Methodology

This publication is intended to discover recent ADR stories, evaluate relevance, merge overlapping reports, and publish structured summaries with source transparency.

Stories are grouped across languages, rewritten into a fixed editorial format, and linked to original sources. How we report.

Source discovery

The pipeline uses a multilingual feed search layer and then fetches source pages only for approved items.

Candidate stories are filtered first with keyword and topic rules and then with AI classification for relevance.

Story generation

Related items are clustered into one event page, summarized into a canonical story, and translated into configured publication languages.

Source lists remain visible so readers can verify reporting origins directly.