Enums

DatasetType

class src.enums.datasets.DatasetType(value)

Bases: str, Enum

This enum is used to specify what type of dataset to use,

DYNAMIC: str = 'dynamic_general'

MULTI_SECOND_LEVEL_ALL_BASED: str = 'm2_general'

MULTI_TOP_LEVEL_ALL_BASED: str = 'm1_general'

MULTI_TOP_LEVEL_ARTICLE_BASED: str = 'm1_article'

MULTI_TOP_LEVEL_ARTICLE_SPLIT: str = 'm1_article_split'

MULTI_TOP_LEVEL_DESCRIPTION_BASED: str = 'm1_description'

MULTI_TOP_LEVEL_MOTIVATION_BASED: str = 'm1_motivation'

MULTI_TOP_LEVEL_SHORT_TITLE_BASED: str = 'm1_shorttitle'

SINGLE_BASIC: str = ''

SINGLE_TOP_LEVEL_ALL_BASED: str = 's1_general'

SUMMARY_STATISTIC_DATASET: str = 'summary_stat_dataset'

UNPROCESSED: str = 'mirror'

classmethod _list()

internal classmethod that allows us to retrieve all possible datasets :return:

static get_from_prefix(model_type: str)

this function allows us to retrieve only the models compliant with the prefix filter

Parameters:

model_type – the string prefix to filter the models with

Returns:

a list with models that comply with the filter

static get_multilevel_datasets(level: int = 1)

this function allows us to retrieve only the multilabel datasets of a specific level

Parameters:

level – the label level you want to retrieve datasets for

Returns:

a list with dataset that comply with the filter

get_single_level_datasets()

DecisionQuery

class src.enums.decision.DecisionQuery(value)

Bases: str, Enum

This enum is used to identify what type of query to run

ALL = 'all'

ANNOTATED = 'annotated'

static match(config: DataModelConfig, value: DecisionQuery)

this function allows us to verify the provided input and return the correct query

Parameters:

• config – the global configuration object

• value – the enum value that is used.

Returns:

format-able query in string format

Graph

class src.enums.graph.GraphType(value)

Bases: str, Enum

This enum contains the predefined graph suffixes that can be used to save to a certain sparql graph

MODEL_ANNOTATION = 'model_annotation'

MODEL_INFORMATION = 'model_information'

TESTING = 'testing_annotation'

USER_ANNOTATION = 'user_annotation'

static match(config: DataModelConfig, value: GraphType)

this function returns the config value for the chosen enum type :param config: :param value: :return:

ModelType

class src.enums.models.ModelType(value)

Bases: str, Enum

This enum is used to identify what model type/flavour you are running.

DYNAMIC_TOPIC_MODEL: str = 'topic_model_dynamic'

EMBEDDING_CHILD_LABELS: str = '_embedding_child_labels'

EMBEDDING_CHUNKED: str = 'embedding_chunked'

EMBEDDING_GROUND_UP: str = '_embedding_ground_up'

EMBEDDING_GROUND_UP_GREEDY: str = '_embedding_ground_up_greedy'

EMBEDDING_REGULAR: str = 'embedding_regular'

EMBEDDING_SENTENCE: str = 'embedding_sentence'

HIERARCHIC_TOPIC_MODEL: str = 'topic_model_hierarchic'

HUGGINGFACE_MODEL: str = 'huggingface_model'

HYBRID_BASE_MODEL: str = 'hybrid_base_model'

HYBRID_SELECTIVE_MODEL: str = 'hybrid_selective_model'

REGULAR_TOPIC_MODEL: str = 'topic_model_regular'

ZEROSHOT_CHILD_LABELS: str = '_zeroshot_child_labels'

ZEROSHOT_CHUNKED: str = 'zeroshot_chunked'

ZEROSHOT_REGULAR: str = 'zeroshot_regular'

ZEROSHOT_SENTENCE: str = 'zeroshot_sentence'

classmethod _list()

internal classmethod that allows us to retrieve all possible datasets :return:

static get_from_prefix(model_type: str)

this function allows us to retrieve only the models compliant with the prefix filter

Parameters:

model_type – the string prefix to filter the models with

Returns:

a list with models that comply with the filter

static get_models_for_type(model_type: str)

this function allows us to retrieve only the models compliant with the prefix filter

Parameters:

model_type – the string prefix to filter the models with

Returns:

a list with models that comply with the filter

EndpointType

class src.enums.request.AuthType(value)

Bases: str, Enum

This enum specifies what authentication type to use on the sparql endpoint

BASIC = 'basic'

DIGEST = 'digest'

NONE = 'none'

class src.enums.request.EndpointType(value)

Bases: str, Enum

This enum is used to specify what type of query you want to execute

DECISION = 'decision'

TAXONOMY = 'taxonomy'

static match(config: Config, value: EndpointType)

this function allows us to verify the provided input and return the correct query

Parameters:

• config – the global configuration object

• value – the enum value that is used.

Returns:

the chosen endpoint url

SetfitClassifierHeads

class src.enums.setfit.SetfitClassifierHeads(value)

Bases: str, Enum

This enum is used to identify what type of setfit head you want to use

DIFFERENTIABLE_MULTI_OUTPUT: str = 'differentiable_multi-output'

DIFFERENTIABLE_ONE_VS_REST: str = 'differentiable_one-vs-rest'

SKLEARN_CLASSIFIER_CHAIN: str = 'sklearn_classifier-chain'

SKLEARN_MULTI_OUTPUT: str = 'sklearn_multi-output'

SKLEARN_ONE_VS_REST: str = 'sklearn_one-vs-rest'

static get_prefixed_heads(prefix: str) → list[str]

This function filters the setfit heads on values and returns the resulting heads

Parameters:

prefix – string to check if in head value

Returns:

filtered output as list

classmethod list()

static match(config: Config, value: SetfitClassifierHeads)

this function allows us to verify the provided input and return the correct query

Parameters:

• config – the global configuration object

• value – the enum value that is used.

Returns:

format-able query in string format

TrainingFlavours

class src.enums.supervised_flavours.TrainingFlavours(value)

Bases: str, Enum

This enum is used to identify what type of training flavour you want to use.

BERT: str = 'bert'

DISTIL_BERT: str = 'distil_bert'

SETFIT: str = 'setfit'

static get_default_model_for_type(config: Config, model_flavour: TrainingFlavours) → str

this function checks what the chosen flavour is and returns the defaulted value from the config for the model_id

Parameters:

• config – the global configuration object

• model_flavour – the enum value that is used.

Returns:

format-able query in string format

TaxonomyFindTypes

class src.enums.taxonomy.TaxonomyFindTypes(value)

Bases: str, Enum

This enum is used to identify what type of taxonomy find method you want to use.

LABEL: str = 'label'

URI: str = 'uri'