Article Identity Collisions in a Focused Publishing Library

Research question: Which identity fields best detect near-duplicate articles before publication?

Method: Construct sixty pairs covering exact matches, revised slugs, similar titles, mirrored outlines, related Blog and Research pages and distinct controls; compare three blinded review cards. Preserve every failed, unavailable and ambiguous observation in the denominator, and record who reviewed it and when.

Analysis should separate prevention from detection and technical conformance from editorial quality. Report results by defect type rather than reducing unlike failures to one score.

Scope and limitations: The reference classification depends on editorial judgment and does not show how search engines evaluate similarity. Synthetic cases, a narrow Philippines employment-operations vocabulary and one application architecture limit generalization. The study cannot establish legal obligations, search performance or provider quality.

The evidence can inform a publishing control, but the authorized editor retains claim and release decisions. A useful conclusion states what was observed, what remains unknown and which owner must resolve exceptions.

Sources consulted: Google Search Central, Article structured data (https://developers.google.com/search/docs/appearance/structured-data/article); Google Search Central, canonical URLs (https://developers.google.com/search/docs/crawling-indexing/consolidate-duplicate-urls); Sitemaps protocol (https://www.sitemaps.org/protocol.html); NIST Cybersecurity Framework 2.0 (https://www.nist.gov/cyberframework). These sources provide general web and control concepts. They do not measure this site, guarantee indexing, certify content quality, or replace editorial, legal, or employment judgment.