nDCG@10
A ranking quality evaluation metric commonly used in information retrieval. It measures how closely the top 10 results returned by a retriever approach an ideal ranking, with values from 0 to 1.
Timeline (last 90 days)
In stress tests with isolated targets, just five query-targeted documents per query can reduce the nDCG@10 of classic retrievers by 40.2% to 61.5%.
nDCG@10 is normalized Discounted Cumulative Gain, which measures how close the top 10 retrieval results are to the ideal ranking, ranging from 0 to 1, with higher values indicating that the top-ranked results are closer to the ideal state.
On the nDCG@10 ranking metric, SAFE-QE shows no statistically significant difference from generative expansion and corpus-guided expansion.
All Facts (3)
In stress tests with isolated targets, just five query-targeted documents per query can reduce the nDCG@10 of classic retrievers by 40.2% to 61.5%.
50%UnverifiednDCG@10 is normalized Discounted Cumulative Gain, which measures how close the top 10 retrieval results are to the ideal ranking, ranging from 0 to 1, with higher values indicating that the top-ranked results are closer to the ideal state.
50%UnverifiedOn the nDCG@10 ranking metric, SAFE-QE shows no statistically significant difference from generative expansion and corpus-guided expansion.
50%