Contextual[source]

The eds.contextual_qualifier pipeline component qualifies spans based on contextual information.

Parameters

PARAMETER DESCRIPTION

nlp

The spaCy pipeline object.

TYPE: PipelineProtocol DEFAULT: None

name

The name of the component.

TYPE: Optional[str] DEFAULT: "contextual_qualifier"

span_getter

The function or callable to get spans from the document.

TYPE: SpanGetterArg

patterns

A dictionary of patterns to match in the text. Each pattern dictionary should follow the structure of the ClassPatternsContext data class.

ClassPatternsContext

PARAMETER	DESCRIPTION
`terms`	Terms to match. TYPE: `Optional[Patterns]` DEFAULT: `None`
`regex`	Regular expressions to match. TYPE: `Optional[Patterns]` DEFAULT: `None`
`context_words`	Number of words to consider as context. TYPE: `Union[NonNegativeInt, Tuple[NonNegativeInt, NonNegativeInt]]` DEFAULT: `0`
`context_sents`	Number of sentences to consider as context. TYPE: `Union[NonNegativeInt, Tuple[NonNegativeInt, NonNegativeInt]]` DEFAULT: `1`
`attr`	Attribute to match on. TYPE: `str` DEFAULT: `'TEXT'`
`regex_flags`	Flags for regular expressions. TYPE: `Union[RegexFlag, int]` DEFAULT: `0`
`ignore_excluded`	Whether to ignore excluded tokens. TYPE: `bool` DEFAULT: `False`
`ignore_space_tokens`	Whether to ignore space tokens. TYPE: `bool` DEFAULT: `False`

TYPE: Dict[str, Dict[Union[str, int], ClassPatternsContext]]