Content classification evidence lab
Bible BookLooks Evaluation
A passage-level content classification hub for loaded Bible translations, using the same content-based rating framework so versions can be inspected and compared.
The point
Classification, not theological judgment.
This project applies a content-based rating framework to passages across the loaded NIV, KJV, and ESV evaluations. The purpose is classification and review, not theological endorsement or rejection.
The central question is whether the same standard is applied consistently and whether readers can inspect the evidence trail themselves.
Rating system
Levels 0 through 5
Findings at a glance
Generated from parsed evidence records
Evidence records
438Parsed passage recordsBooks with Level 2+ triggers
55Generated by bookBooks with no 2+ triggers
11Preserved in metadataLevel 5 records
22Highest severity classBrowse the evidence
Inspect classifications passage by passage.
Search references, books, categories, and rationale text. Policy-dependent and wording-sensitive records are labeled instead of buried.
Translation comparison
Same references, multiple translations.
NIV, KJV, and ESV are loaded against the same 438 canonical passage records. The comparison view flags wording-sensitive rationale changes while preserving the shared rating standard.
Additional translations can be added later against the same canonical passage keys.