Knowledge Extractor

The knowledge extractor will perform at the same time the extraction and classification of named entities and the extraction of relations among them.

Currently, the are only two Knowledge Extractor available: KnowGL and ReLiK

Bases: ABC

Instantiate the Knowledge Extractor

Parameters:

Name	Type	Description	Default
device	Optional[Union[str, device]]	Device to be used for computation	None

__hash__()

Get hash representation of the component

extract_knowledge(docs, batch_size=None)

Perform the relations extraction. Call the predict function and add the mentions to the Spacy Doc

Parameters:

Name	Type	Description	Default
docs	Iterator[Doc]	A list of spacy Document	required
batch_size	Optional[Union[int, None]]	The batch size	None

Returns:

Type	Description

from_disk(path, exclude=()) classmethod

Load component from disk

load_models()

Load the model

Returns:

Type	Description

parse_triples(preds)

Parse the triples into lists of entities and relations

Parameters:

Name	Type	Description	Default
preds	List[Tuple[Span, RelationSpan, Span]]	Predicted triples	required

Returns:

Type	Description
Tuple[List[Span], List[RelationSpan]]	Tuple with list of entities and list of relations

predict(docs, batch_size=None) abstractmethod

Perform the knowledge extraction.

Parameters:

Name	Type	Description	Default
docs	Iterator[Doc]	A list of spacy Document	required
batch_size	Optional[Union[int, None]]	The batch size	None

Returns:

Type	Description
List[List[Tuple[Span, RelationSpan, Span]]]	the predicted triples

set_device(device)

Set the device to use

Parameters:

Name	Type	Description	Default
device	Union[str, device]		required

Returns:

Type	Description

to_disk(path)

Save component into disk