dialoguekit.nlu.intent_classifier

Abstract interface for intent classification.

This interface assumes a single intent per utterance, i.a., approaches the task as a single-label classification problem. The generalization to multi-label classification is left to future work.

Classes

IntentClassifier

Initializes the intent classifier.

Module Contents

class dialoguekit.nlu.intent_classifier.IntentClassifier(intents: List[dialoguekit.core.intent.Intent])

Bases: abc.ABC

Initializes the intent classifier.

Parameters:

intents – List of allowed intents.

abstract train_model(utterances: List[dialoguekit.core.utterance.Utterance], labels: List[dialoguekit.core.intent.Intent]) → None

Trains a model based on a set of labeled utterances.

Parameters:

• utterances – List of Utterance instances.

• labels – List of associated intent labels.

Raises:

NotImplementedError – If not implemented in derived class.

abstract classify_intent(utterance: dialoguekit.core.utterance.Utterance) → dialoguekit.core.intent.Intent

Classifies the intent of an utterance.

Parameters:

utterance – An utterance.

Returns:

Predicted intent.

Raises:

NotImplementedError – If not implemented in derived class.

abstract save_model(file_path: str) → None

Saves the trained model to a file.

Parameters:

file_path – File path.

Raises:

NotImplementedError – If not implemented in derived class.

abstract load_model(file_path: str) → None

Loads a model from a file.

Parameters:

file_path – File path.

Raises:

NotImplementedError – If not implemented in derived class.