framenet_tools.fee_identification package


framenet_tools.fee_identification.feeidentifier module

class framenet_tools.fee_identification.feeidentifier.FeeIdentifier(cM: framenet_tools.config.ConfigManager)

Bases: object

evaluate_acc(dataset: List[List[str]])

Evaluates the accuracy of the Frame Evoking Element Identifier

NOTE: F1-Score is a better way to evaluate the Identifier, because it tends to predict too many FEEs

Parameters:dataset – The dataset to evaluate
Returns:A Triple of the count of correct elements, total elements and the accuracy
identify_targets(sentence: list)

Identifies targets for a given sentence

Parameters:sentence – A list of words in a sentence
Returns:A list of targets
predict_fees(mReader: framenet_tools.data_handler.reader.DataReader)

Predicts the Frame Evoking Elements NOTE: This drops current annotation data

predict_fees_old(dataset: List[List[str]])

Predicts all FEEs for a complete datset

Parameters:dataset – The dataset to predict
Returns:A list of predictions
query(x: List[str])

Query a prediction of FEEs for a given sentence

Parameters:x – A list of words in a sentence
Returns:A list of predicted FEEs
framenet_tools.fee_identification.feeidentifier.should_include_token(p_data: list)

A static syntactical prediction of possible Frame Evoking Elements

Parameters:p_data – A list of lists containing token, pos_tag, lemma and NE
Returns:A list of possible FEEs

Module contents