Study Notes

Home

❯

Concepts

❯

MRR

MRR

Jun 06, 20261 min read

  • evaluation
  • key-formula

Mean Reciprocal Rank (MRR)

MRR

MRR=∣Q∣1​∑i=1∣Q∣​ranki​1​

where ranki​ is the rank of the first relevant document for query i.

Useful when you care about how quickly the user finds any relevant result (e.g., question answering, navigational queries).

Appears In

  • IR-L04 - Evaluation
  • RS-L01 - Course Overview & Introduction
  • RS-L02 - Evaluation Beyond Accuracy

Graph View

  • Mean Reciprocal Rank (MRR)
  • Appears In

Backlinks

  • B Testing
  • BERT4Rec
  • Beyond-Accuracy Metrics
  • Collaborative Filtering
  • Content-Based Recommendation
  • Diversity
  • GRU4Rec
  • Hit Rate
  • Implicit and Explicit Feedback
  • LambdaMART
  • Long-Tail Distribution
  • Matrix Factorization
  • Neighborhood-based Collaborative Filtering
  • Online and Offline Evaluation
  • Recommender System
  • Serendipity
  • Session-based Recommendation
  • Top-N Recommendation
  • IR-PTR Ch2 - Setting the Stage
  • IR-A01 - Unsupervised Retrieval
  • IR - Overview
  • IR-L04 - Evaluation
  • RS-L01 - Course Overview & Introduction
  • RS-L02 - Evaluation Beyond Accuracy
  • RecSys - Overview

Created with Quartz v4.5.2 © 2026

  • GitHub
  • Discord Community