The Rank Factor Class taxonomy categorises content hypotheses into six structural classes—alignment, substance, architecture, style, framing, and proof—to prevent overlapping tests and maintain exploration diversity across Bayesian optimization rounds.
flowchart TD
A[Rank Factor Classes] --> C1[1. Alignment]
A --> C2[2. Substance]
A --> C3[3. Architecture]
A --> C4[4. Style]
A --> C5[5. Framing]
A --> C6[6. Proof]
C1 --> F1[Token and Intent Match]
C2 --> F2[Entities and Quantification]
C3 --> F3[Positioning and Format]
C4 --> F4[Tone and Conciseness]
C5 --> F5[Perspective and Angle]
C6 --> F6[Citations and Evidence]Referenced by