get_noun_phrases

parse_utils.get_noun_phrases(
    corp,
    nlp_model,
    n_process=1,
    batch_size=25,
    disable_ner=True,
)

Extract expanded noun phrases using the ‘en_core_web_sm’ model.

Parameters

Name	Type	Description	Default
corp	pl.DataFrame	A polars DataFrame conataining a ‘doc_id’ column and a ‘text’ column.	required
nlp_model	Language	An ‘en_core_web_sm’ instance.	required
n_process		The number of parallel processes to use during parsing.	1
batch_size		The batch size to use during parsing.	25

Returns

Name	Type	Description
	pl.DataFrame	a polars DataFrame with, noun phrases and their assocated part-of-speech tags.

Notes

Noun phrases can be extracted directly from the noun_chunks attribute. However, per spaCy’s documentation the attribute does not permit nested noun phrases, for example when a prepositional phrases modifies a preceding noun phrase. This function extracts elatorated noun phrases in their complete form.