How should I deal with duplicated results?¶
Group likely copies in the analysis, choose a primary artefact and retain any additional trace that contributes provenance, state or context.
Decide whether they represent one event¶
Compare event or message IDs, participants, direction, time, content, account, attachment and source path. Similar wording may be a resend or edit; a notification may be only a preview. Several parser categories pointing to the same source row are not independent confirmations.
Preserve meaningful differences¶
Prefer the record with complete content, source and status as primary. A journal, notification, backup or recipient device may establish an earlier state, display or independent copy. Identical file hashes establish content identity, not identical conversational or device context. Document the grouping rather than deleting source results.
Key takeaway
Deduplicate the explanation, not the evidence trail, and state why records were grouped or retained separately.