How Reliably Can Reviewers Link Duplicate HR Messages?

Research question: Can a minimal field set link repeated HR messages without merging distinct sensitive matters or overlooking new evidence?

Methodology and scope: Build eighty synthetic pairs: exact resends, cross-channel repeats, amended requests, unrelated matters, similar subjects, forwarded copies, and ambiguous pairs. Pre-label link, separate, or escalate under a written definition.

Protocol and measures: Two reviewers use sender token, category, dates, case reference, attachment fingerprint, and stated change. Report false merges, missed links, escalation rate, agreement, and time per pair.

Scope boundary: The study evaluates linkage, not merit or urgency. Reviewers see redacted fields and cannot open free-text attachments. It does not decide access, retention, response, or employment outcomes.

Inference boundaries: Low false-merge rates would support a limited human-reviewed pilot, not automated matching or shared access. HR and privacy owners decide consolidation, viewers, response, retention, and exceptions.

Limitations: Synthetic language lacks live ambiguity, emotion, multilingual phrasing, and identity uncertainty. Seeded prevalence affects predictive values. A pilot needs privacy review, logging, escalation, and a mislink stop rule.

References: National Institute of Standards and Technology, Cybersecurity Framework 2.0 (https://doi.org/10.6028/NIST.CSWP.29); National Institute of Standards and Technology, Privacy Framework 1.0 (https://doi.org/10.6028/NIST.CSWP.01162020); U.S. Government Accountability Office, Standards for Internal Control in the Federal Government, GAO-14-704G (https://www.gao.gov/products/gao-14-704g); Philippine Statistics Authority, Annual Survey of Philippine Business and Industry (https://psa.gov.ph/statistics/annual-survey-philippine-business-and-industry). These primary publications supply control, privacy, evidence, and sector context. None evaluates Outsourced Employment or the synthetic cases proposed here.