Highest value of machine translation metrics Does anyone knows the highest and lowest values of the common machine translation metrics $BLEU_{1-4}$, $METEOR$, $ROUGE_{L}$, and $CIDEr$ and $CIDEr-D$?
I know that for BLEU is 1, correct? Does the same stands for all other metrics mentioned above? 
Also, what about $CIDEr$ and $CIDEr-D$ metric? What is the maximum value? 
 A: *

*$BLEU_{1-n}$: minimum 0, maximum 1.

*$METEOR$: minimum 0, maximum 1.

*$ROUGE_{n}$, $ROUGE_{L}$, $ROUGE_{S}$: minimum 0, maximum 1.

*$CIDEr$: minimum 0, maximum 1. Original reference: {1}.

*$CIDEr-D$?: minimum 0, maximum 1.  $CIDEr-D$ = CIDEr Defended, named as such as it is more robust to gaming effects. Original reference: {1}.


Keep in mind that sometime people uses percentage. E.g., one could say $ROUGE_{n}=40\%$.
{2} goes over all these metrics in the context of the image captioning evaluation. Note that for ROUGE there are some small differences between how {2} (and many other papers) present ROUGE and the original ROUGE implementation (e.g., it does have stemming).

References:


*

*{1} Vedantam, Ramakrishna, C. Lawrence Zitnick, and Devi Parikh. "CIDEr: Consensus-based image description evaluation." In Proceedings of the IEEE conference on computer vision and pattern recognition, pp. 4566-4575. 2015. https://scholar.google.com/scholar?cluster=2428149707553201565&hl=en&as_sdt=0,5 ; https://arxiv.org/abs/1411.5726

*{2} Chen, Xinlei, Hao Fang, Tsung-Yi Lin, Ramakrishna Vedantam, Saurabh Gupta, Piotr Dollár, and C. Lawrence Zitnick. "Microsoft COCO captions: Data collection and evaluation server." arXiv preprint arXiv:1504.00325 (2015).;  https://scholar.google.com/scholar?cluster=4232572309886474756&hl=en&as_sdt=0,5 ; https://arxiv.org/abs/1504.00325
