Datasets:
GEM
/

Tasks:
Other
Multilinguality:
unknown
Size Categories:
unknown
Language Creators:
unknown
Annotations Creators:
expert-created
Source Datasets:
original
License:
mathiascreutz commited on
Commit
6306358
1 Parent(s): 6ca6aaf

Minor modifications

Browse files
Files changed (1) hide show
  1. README.md +35 -0
README.md CHANGED
@@ -130,6 +130,41 @@ TBA
130
 
131
  `gem_id`: unique identifier of this entry
132
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
133
  ### Data Splits
134
 
135
  The data is split into ...
 
130
 
131
  `gem_id`: unique identifier of this entry
132
 
133
+ **Additional information about the annotation scheme:**
134
+
135
+ The annotation scores given by an individual annotator are:
136
+
137
+ 4: Good example of paraphrases = Dark green button in the annotation
138
+ tool = The two sentences can be used in the same situation and
139
+ essentially "mean the same thing".
140
+
141
+ 3: Mostly good example of paraphrases = Light green button in the
142
+ annotation tool = It is acceptable to think that the two sentences
143
+ refer to the same thing, although one sentence might be more specific
144
+ than the other one, or there are differences in style, such as polite
145
+ form versus familiar form.
146
+
147
+ 2: Mostly bad example of paraphrases = Yellow button in the annotation
148
+ tool = There is some connection between the sentences that explains
149
+ why they occur together, but one would not really consider them to
150
+ mean the same thing.
151
+
152
+ 1: Bad example of paraphrases = Red button in the annotation tool =
153
+ There is no obvious connection. The sentences mean different things.
154
+
155
+ If the two annotators fully agreed on the category, the value in the
156
+ `annot_score` field is 4.0, 3.0, 2.0 or 1.0. If the two annotators
157
+ chose adjacent categories, the value in this field will be 3.5, 2.5 or
158
+ 1.5. For instance, a value of 2.5 means that one annotator gave a
159
+ score of 3 ("mostly good"), indicating a possible paraphrase pair,
160
+ whereas the other annotator scored this as a 2 ("mostly bad"), that
161
+ is, unlikely to be a paraphrase pair. If the annotators disagreed by
162
+ more than one category, the sentence pair was discarded and won't show
163
+ up in the datasets.
164
+
165
+ The training sets were not annotated manually. This is indicated by
166
+ the value 0.0 in the `annot_score` field.
167
+
168
  ### Data Splits
169
 
170
  The data is split into ...