yogsoth-ai Category Sorting
Purpose: Classify candidate alternatives into predefined categories (e.g., A/B/C grades, compliant/non-compliant), supporting ELECTRE-Tri, FlowSort, AHPSort, DRSA, and other classification methods.
When to use:
- User needs to classify alternatives rather than rank them
- Predefined category boundaries exist (e.g., pass/fail, excellent/good/poor)
- Need to independently determine category membership for each alternative
Budget
| Base SOP | Target | ±10% Range |
|---|---|---|
| criterion-definition | 5-8 criteria | 4-9 |
| weight-elicitation-sop | 1 weight vector | 1 |
| threshold-setting | 1 threshold set | 1 |
| alternative-scoring | 1 score matrix | 1 |
| scoring-synthesis | 1 classification | 1 |
State Ledger
strategy: category-sorting
status: pending
categories_defined: false
criteria_defined: false
weights_computed: false
thresholds_set: false
scores_computed: false
classified: false
result: null
Available Tactics
- scoring-matrix-construction — Build scoring foundation
- screening-then-scoring — Hybrid workflow: screen first, then classify
Available SOPs
Import (from tactics)
- criterion-definition
- weight-elicitation-sop
- alternative-scoring
- threshold-setting
Subagent
- scoring-synthesis
Execution Guidance
- Define category definitions and boundary conditions
- Invoke criterion-definition to determine classification criteria
- Invoke threshold-setting to set category boundaries
- Score each alternative independently and determine category membership
- Handle borderline cases (pessimistic vs optimistic assignment)
Output Format
## Classification Results
**Method:** [ELECTRE-Tri / FlowSort / AHPSort / DRSA]
**Category Definitions:** [A=Excellent, B=Good, C=Needs Improvement, D=Unqualified]
### Classification Table
| Alternative | Category | Confidence | Boundary Distance |
|-------------|----------|------------|-------------------|
### Borderline Cases
[List alternatives near category boundaries and their sensitivity]
Available Tactics
Optional, no fixed order; the final leaf is always a sop.
| Tactic | When to use |
|---|---|
| convergence-scoring-matrix-construction | Build a complete scoring matrix through criterion definition, weighting, scoring, normalization, and sensitivity testing. |
| screening-then-scoring | First eliminate non-qualifying candidates with non-compensatory rules, then score survivors with full MCDA methods. |