Learning to Rank by Optimizing NDCG Measure

By Hamed Valizadegan, Rong Jin, Ruofei Zhang, Jianchang Mao. 2009

228 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

1. Tie-Yan Liu
2. Yashoteja Prabhu
3. Liwei Wang
4. Dacheng Tao
5. Manik Varma
6. Di He
7. Xu Tan
8. Taiji Suzuki

Citations in the corpus

• BLENDER: Enabling Local Search with a Hybrid Differential Privacy Model (2017, background): “NDCG: NDCG is a standard measure of search quality [20, 38] that explicitly takes the ordering of the items in a results list into account.”
• Cross-domain Recommendation via Deep Domain Adaptation (2018, methodology): “Third, as optimizing the model with nDCG saw a large improvement, we also suggest replacing the softmax loss with another ranking loss such as nDCG [33].”
• A Theoretical Analysis of NDCG Type Ranking Measures (2013, background): “Although early results of learning to rank often reduce ranking problem to classification or regression [15, 18, 23, 26, 5], recently there is evidence that learning a ranking function by optimizing a ranking measure such as NDCG is a promising approach [33, 36].”
• Learning to Rank Lexical Substitutions (2013, methodology): “Finally, in the listwise approach, tailor-made learning methods are used that directly optimize the ranking performance with respect to a global evaluation metric, i.e., a measure that evaluates the ranking of a complete set of query instances (Valizadegan et al., 2009).”
• FastXML: a fast, accurate and stable tree-classifier for extreme multi-label learning (2014, background): “It is well recognized in the learning to rank literature that nDCG is a difficult function to optimize [25,26,31,32] since it is sharply discontinuous with respect to w and hence standard stochastic sub-gradient descent techniques cannot be applied.”
• MLAIRE: Multilingual Language-Aware Information Retrieval Evaluation Protocal (2026, unknown): “When the query-language passage and its non-query-language translations are all semantically relevant, standard metrics such as Recall and nDCG treat them as equally relevant [20, 21].”