lexnlp.extract.ml.en.definitions package¶
Subpackages¶
Submodules¶
lexnlp.extract.ml.en.definitions.definition_phrase_detector module¶
-
class
lexnlp.extract.ml.en.definitions.definition_phrase_detector.
DefinitionPhraseDetector
¶ Bases:
lexnlp.extract.ml.detector.artifact_detector.ArtifactDetector
Search for the phrase surrounding the term being defined
Let the prase be <agrees to serve the Company in such capacity during the term of employment (the “Employment Period”).
… model_definition will find <term of employment (the “Employment Period”)>
-
process_sample
(sample_df: pandas.core.frame.DataFrame, build_target_data: bool = False) → Union[numpy.ndarray, Tuple[numpy.ndarray, numpy.ndarray]]¶
-
train_and_save
(settings: lexnlp.extract.ml.detector.detecting_settings.DetectingSettings, train_file: str, train_size: int = -1, save_path: str = '', compress: bool = False) → None¶ Create a percent identification model using tokens. :param settings: Model settings :param train_file: File to load training samples from :param train_size: Number of records to use :param save_path: Output (pickle model) file path :param compress: Save compressed file
-
train_and_save_on_dataframe
(settings: lexnlp.extract.ml.detector.detecting_settings.DetectingSettings, train_sample_df: pandas.core.frame.DataFrame, save_path: str = '', compress: bool = False) → None¶
-
lexnlp.extract.ml.en.definitions.definition_term_detector module¶
-
class
lexnlp.extract.ml.en.definitions.definition_term_detector.
DefinitionTermDetector
¶ Bases:
lexnlp.extract.ml.detector.artifact_detector.ArtifactDetector
-
process_sample
(sample_df: pandas.core.frame.DataFrame, build_target_data: bool = False) → Union[numpy.ndarray, Tuple[numpy.ndarray, numpy.ndarray]]¶
-
train_and_save
(settings: lexnlp.extract.ml.detector.detecting_settings.DetectingSettings, train_file: str, train_size: int = -1, save_path: str = '', compress: bool = False) → None¶ Create a percent identification model using tokens. :param settings: Model settings :param train_file: File to load training samples from :param train_size: Number of records to use :param save_path: Output (pickle model) file path :param compress: Save compressed file
-
train_and_save_on_dataframe
(settings: lexnlp.extract.ml.detector.detecting_settings.DetectingSettings, train_sample_df: pandas.core.frame.DataFrame, save_path: str = '', compress: bool = False) → None¶
-
lexnlp.extract.ml.en.definitions.layered_definition_detector module¶
-
class
lexnlp.extract.ml.en.definitions.layered_definition_detector.
LayeredDefinitionDetector
¶ Bases:
object
-
get_annotations
(sentence: str) → List[lexnlp.extract.common.annotations.definition_annotation.DefinitionAnnotation]¶
-
static
join_adjacent_definitions_labels
(labels_definitions, labels_terms, row_text)¶
-
load_compressed
(file_path: str)¶ Loads archive with two model pickle files (model_definition, model_term)
-
train_on_doccano_jsonl
(save_file_path: str, exported_doc_path: str, text_column_name: str = 'text', labels_column_name: str = 'labels', label_term: str = 'term', label_definition: str = 'definition')¶
-
train_on_formatted_data
(definition_frame: pandas.core.frame.DataFrame, term_frame: pandas.core.frame.DataFrame, save_file_path: str)¶ Parameters: - definition_frame – dataframe, [ (row_text, [(start, end), (start, end)…], feature_mask]
- term_frame – dataframe, [ (row_text, [(start, end), (start, end)…]]
- save_file_path – path to store zipped model files (as one file)
-