Where each duplicate-detection method sits

Positioned by what a method can catch (literal spelling → conceptual meaning) against how much infrastructure and cost it takes to run. Hover or tab through a point for details.

↗ cost & infrastructure climb as a method understands more meaning best catch-rate per dollar for most real-world duplicates Low High Infrastructure & cost → Literal Conceptual What it catches →
String-based methods (spelling & structure)
Semantic / vector-based methods (meaning)

Axes are directional, not measured units — the map shows relative position, not precise scores.