119
click to vote
SIGIR
2006
ACM
15 years 10 months ago
2006
ACM
We introduce and validate bootstrap techniques to compute confidence intervals that quantify the effect of test-collection variability on average precision (AP) and mean average...