mindmeld.models.entity_features module

This module contains feature extractors for entities

mindmeld.models.entity_features.extract_bag_of_words_after_features(ngram_lengths_to_start_positions, **kwargs)[source]

Returns a bag-of-words feature extractor.

Parameters:ngram_lengths_to_start_positions (dict) --
Returns:(function) The feature extractor.
mindmeld.models.entity_features.extract_bag_of_words_before_features(ngram_lengths_to_start_positions, **kwargs)[source]

Returns a bag-of-words feature extractor.

Parameters:ngram_lengths_to_start_positions (dict) --
Returns:(function) The feature extractor.
mindmeld.models.entity_features.extract_in_gaz_features(**kwargs)[source]

Returns a feature extractor that finds any gazetteer matches against the input query

mindmeld.models.entity_features.extract_numeric_candidate_features(**kwargs)[source]

Returns a feature extractor that generates features indicating the presence of the sys_time and sys_interval numeric entities. These numeric entities are identified by duckling, the numerical entity recognition service and boosted by training data containing the entity labels. Used by the role classifier when the 'numeric' feature is specified in the config.

mindmeld.models.entity_features.extract_other_entities_features(**kwargs)[source]

Returns a feature extractor for all other entities apart from the current entity. Used by the role classifier when the 'other-entities' feature is specified in the config.