# extr-ds
> Library to programmatically build labeled datasets for Named-Entity Recognition (NER) and Relation Extraction (RE) Machine Learning tasks.
<br />
## Install
```
pip install extr-ds
```
## Command Line
see [Instructions](https://medium.com/@pasdan/building-custom-named-entity-recognition-models-e4d8d95804e) on how to use the command line utility to manage your project.
### 1. Init Project
```
extr-ds --init
```
### 2. Split and Annotate
```
extr-ds --split
```
### 3.a Annotate Entities or Relations Again?
```
extr-ds --annotate -ents
extr-ds --annotate -rels
```
### 3.b Change Relation Extraction Label
```
extr-ds --relate -label NO_RELATION=5,7,9
```
### 3.b Remove Relation Extraction Instance
```
extr-ds --relate -delete 5,6,7
```
### 4. Save
```
extr-ds --save -ents
extr-ds --save -rels
```
## API
## Example
```python
text = 'Ted Johnson is a pitcher. Ted went to my school.'
```
### 1. Label Entities for Named-Entity Recognition Task (NER)
```python
from extr import RegEx, RegExLabel
from extr.entities import EntityExtactor
from extr_ds.labelers import IOB
entity_extractor = EntityExtactor([
RegExLabel('PERSON', [
RegEx([r'(ted\s+johnson|ted)'], re.IGNORECASE)
]),
RegExLabel('POSITION', [
RegEx([r'pitcher'], re.IGNORECASE)
]),
])
sentence_tokenizer = ## 3rd party tokenizer ##
labels = IOB(sentence_tokenizer, entity_extractor).label(text)
## labels == [
## <Label tokens=..., labels=['B-PERSON', 'I-PERSON', 'O', 'O', 'B-POSITION', 'O']>,
## <Label tokens=..., labels=['B-PERSON', 'O', 'O', 'O', 'O', 'O']>
## ]
```
### 2. Annotate for Relation Extraction Task (RE)
```python
from extr.entities import EntityExtractor
from extr.relations import RegExRelationLabelBuilder, \
RelationExtractor
from extr_ds.labelers import RelationClassification
person_to_position_relationship = RegExRelationLabelBuilder('is_a') \
.add_e1_to_e2(
'PERSON',
[
r'\s+is\s+a\s+',
],
'POSITION'
) \
.build()
labeler = RelationClassification(
sentence_tokenizer,
EntityExtractor([
RegExLabel('PERSON', [
RegEx([r'(ted johnson|bob)'], re.IGNORECASE)
]),
RegExLabel('POSITION', [
RegEx([r'pitcher'], re.IGNORECASE)
]),
]),
RelationExtractor([person_to_position_relationship]),
[('PERSON', 'POSITION', 'NO_RELATION')],
)
labels = labeler.label(text)
## labels == [
## <RelationLabel sentence="<e1>Ted Johnson</e1> is a <e2>pitcher</e2>." label="is_a">
## ]
```
### 3. Find and define the type of difference between labels
```python
from extr_ds.validators import check_for_differences
differences_in_labels = check_for_differences(
['B-PERSON', 'I-PERSON', 'O', 'O', 'B-POSITION', 'O'],
['B-PERSON', 'O', 'O', 'O', 'B-POSITION', 'O']
)
## differences_in_labels.has_diffs == True
## differences_in_labels.diffs_between_labels == [
## <Difference index=1, diff_type=DifferenceTypes.S2_MISSING>
## ]
differences_in_labels = check_for_differences(
['B-PERSON', 'I-PERSON', 'O', 'O', 'B-POSITION', 'O'],
['B-PERSON', 'B-PERSON', 'O', 'O', 'B-POSITION', 'O']
)
## differences_in_labels.has_diffs == True
## differences_in_labels.diffs_between_labels == [
## <Difference index=1, diff_type=DifferenceTypes.MISMATCH>
## ]
```
