Untangling Relevance Score Boosting and Decay Functions - Qdrant

What this post added

This post introduces and explains Qdrant's Linear, Exponential, and Gaussian decay functions for score boosting. It details the purpose of these functions in transforming numeric properties into relevance scores, defines the parameters (x, target, scale, midpoint) and their roles, and provides practical examples for their application. It also addresses the limitation of not being able to dynamically set decay parameters for score normalization when input ranges are unknown.