rbawden commited on
Commit
dcd0d9e
1 Parent(s): e943dc9

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +10 -19
README.md CHANGED
@@ -30,19 +30,18 @@ For machine translation to tackle discourse phenomena, models must have access t
30
 
31
  ## Uses
32
 
33
- <!-- Address questions around how the dataset is intended to be used. -->
 
 
 
 
34
 
35
- ### Direct Use
36
-
37
- <!-- This section describes suitable use cases for the dataset. -->
38
-
39
- [More Information Needed]
40
 
41
  ### Out-of-Scope Use
42
 
43
- <!-- This section addresses misuse, malicious use, and uses that the dataset will not work well for. -->
44
-
45
- [More Information Needed]
46
 
47
  ## Dataset Structure
48
 
@@ -78,15 +77,7 @@ There is no personal or sensitive information in the dataset.
78
 
79
  [More Information Needed]
80
 
81
- ### Recommendations
82
-
83
- <!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
84
-
85
- Users should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.
86
-
87
- ## Citation [optional]
88
-
89
- <!-- If there is a paper or blog post introducing the dataset, the APA and Bibtex information for that should go in this section. -->
90
 
91
  **BibTeX:**
92
 
@@ -113,4 +104,4 @@ Bawden, R., Sennrich, R., Birch, A., & Haddow, B. (2018). Evaluating Discourse P
113
 
114
  ## Dataset Card Authors
115
 
116
- Rachel Bawden= rachel.bawden@inria.fr
 
30
 
31
  ## Uses
32
 
33
+ The dataset is a contrastive evaluation test used to test the use of previous linguistic context (one previous sentence) when translating
34
+ from English to French. Each example block is composed of a previous source sentence and two possible translations, accompanied with
35
+ accompanying context, which can be used to disambiguate the two possible translations. Each instance in the dataset has a correct translation
36
+ and a contrastive one. MT models should be used to rank each translation given the context provided, and the model is evaluated on its
37
+ accuracy of ranking the correct translations above the contrastive ones.
38
 
 
 
 
 
 
39
 
40
  ### Out-of-Scope Use
41
 
42
+ The test set is designed for translation from English to French. The test set cannot be used in the opposite direction, as the ambiguity
43
+ of the source sentence will not hold. The test set is used to be used as a complement to global metrics of evaluation quality and assumes
44
+ that the model has been designed for translation.
45
 
46
  ## Dataset Structure
47
 
 
77
 
78
  [More Information Needed]
79
 
80
+ ## Citation
 
 
 
 
 
 
 
 
81
 
82
  **BibTeX:**
83
 
 
104
 
105
  ## Dataset Card Authors
106
 
107
+ Rachel Bawden: rachel.bawden@inria.fr