Thomas Lemberger commited on
Commit
06fb01f
1 Parent(s): 63f14b6

datset card

Browse files
Files changed (1) hide show
  1. README.md +17 -15
README.md CHANGED
@@ -45,7 +45,9 @@ YAML tags:
45
 
46
  ### Dataset Summary
47
 
48
- This dataset is based on the content of the SourceData (https://sourcedata.embo.org) database, which contains manually annotated figure legends written in English and extracted from scientific papers in the domain of cell and molecular biology (Liechti et al, Nature Methods, 2017, https://doi.org/10.1038/nmeth.4471). The dataset was built to train models for the automatic extraction of a knowledge graph based from the scientific literature. The dataset can be used to train models for text segmentation, named entity recognition and semantic role labeling. The dataset is pre-tokenized with the `roberta-base` tokenizer.
 
 
49
 
50
  ### Supported Tasks and Leaderboards
51
 
@@ -81,13 +83,13 @@ The text in the dataset is English.
81
 
82
  ### Data Fields
83
 
84
- - `tokens`:
85
- - `input_ids`:
86
  - `label_ids`:
87
- - `entity_types`:
88
- - `geneprod_roles`:
89
- - `boring`:
90
- - `panel_start`:
91
 
92
  ### Data Splits
93
 
@@ -106,7 +108,7 @@ The text in the dataset is English.
106
 
107
  ### Curation Rationale
108
 
109
- [More Information Needed]
110
 
111
  ### Source Data
112
 
@@ -116,31 +118,31 @@ The text in the dataset is English.
116
 
117
  #### Who are the source language producers?
118
 
119
- [More Information Needed]
120
 
121
  ### Annotations
122
 
123
  #### Annotation process
124
 
125
- [More Information Needed]
126
 
127
  #### Who are the annotators?
128
 
129
- [More Information Needed]
130
 
131
  ### Personal and Sensitive Information
132
 
133
- [More Information Needed]
134
 
135
  ## Considerations for Using the Data
136
 
137
  ### Social Impact of Dataset
138
 
139
- [More Information Needed]
140
 
141
  ### Discussion of Biases
142
 
143
- [More Information Needed]
144
 
145
  ### Other Known Limitations
146
 
@@ -150,7 +152,7 @@ The text in the dataset is English.
150
 
151
  ### Dataset Curators
152
 
153
- [More Information Needed]
154
 
155
  ### Licensing Information
156
 
 
45
 
46
  ### Dataset Summary
47
 
48
+ This dataset is based on the content of the SourceData (https://sourcedata.embo.org) database, which contains manually annotated figure legends written in English and extracted from scientific papers in the domain of cell and molecular biology (Liechti et al, Nature Methods, 2017, https://doi.org/10.1038/nmeth.4471).
49
+
50
+ The dataset is pre-tokenized with the `roberta-base` tokenizer.
51
 
52
  ### Supported Tasks and Leaderboards
53
 
 
83
 
84
  ### Data Fields
85
 
86
+ - `tokens`: tokens as a`list` of `string` produced by `roberta-base` tokenizer
87
+ - `input_ids`: token id in `roberta-base` tokenizers' vocabulary provided as a`list` of `int`
88
  - `label_ids`:
89
+ - `entity_types`: `list` of `strings` for the IOB2 tags for entity type; possible value in `["O", "I-SMALL_MOLECULE", "B-SMALL_MOLECULE", "I-GENEPROD", "B-GENEPROD", "I-SUBCELLULAR", "B-SUBCELLULAR", "I-CELL", "B-CELL", "I-TISSUE", "B-TISSUE", "I-ORGANISM", "B-ORGANISM", "I-EXP_ASSAY", "B-EXP_ASSAY"]`
90
+ - `geneprod_roles`: `list` of `strings` for the IOB2 tags for experimental roles; values in `["O", "I-CONTROLLED_VAR", "B-CONTROLLED_VAR", "I-MEASURED_VAR", "B-MEASURED_VAR"]`
91
+ - `boring`: `list` of `strings` for IOB2 tags for entities unrelated to causal design; values in ["O", "I-BORING", "B-BORING"]
92
+ - `panel_start`: `list` of `strings` for IOB2 tags `["O", "B-PANEL_START"]`
93
 
94
  ### Data Splits
95
 
 
108
 
109
  ### Curation Rationale
110
 
111
+ The dataset was built to train models for the automatic extraction of a knowledge graph based from the scientific literature. The dataset can be used to train models for text segmentation, named entity recognition and semantic role labeling.
112
 
113
  ### Source Data
114
 
 
118
 
119
  #### Who are the source language producers?
120
 
121
+ The examples are extracted from the figure legends from scientific papers in cell and molecular biology.
122
 
123
  ### Annotations
124
 
125
  #### Annotation process
126
 
127
+ The annotations were produced manually with expert curators from the SourceData project (https://sourcedata.embo.org)
128
 
129
  #### Who are the annotators?
130
 
131
+ Curators of the SourceData project.
132
 
133
  ### Personal and Sensitive Information
134
 
135
+ None known.
136
 
137
  ## Considerations for Using the Data
138
 
139
  ### Social Impact of Dataset
140
 
141
+ Not applicable.
142
 
143
  ### Discussion of Biases
144
 
145
+ The examples are heavily biased towards cell and molecular biology and are enriched in examples from papers published in EMBO Press journals (https://embopress.org)
146
 
147
  ### Other Known Limitations
148
 
 
152
 
153
  ### Dataset Curators
154
 
155
+ Thomas Lemberger, EMBO.
156
 
157
  ### Licensing Information
158