ontolearn.metrics

Quality metrics for concept learners.

Module Contents

Classes

Recall	Recall quality function.
Precision	Precision quality function.
F1	F1-score quality function.
Accuracy	Accuracy quality function.
WeightedAccuracy	WeightedAccuracy quality function.

class ontolearn.metrics.Recall(*args, **kwargs)

Bases: ontolearn.abstracts.AbstractScorer

Recall quality function.

Attribute:

name: name of the metric = ‘Recall’.

__slots__ = ()

name: Final = 'Recall'

score2(tp: int, fn: int, fp: int, tn: int) → Tuple[bool, float]

Quality score for a coverage count.

Parameters:

• tp – True positive count.

• fn – False negative count.

• fp – False positive count.

• tn – True negative count.

Returns:

Tuple, first position indicating if the function could be applied, second position the quality value

in the range 0.0–1.0.

class ontolearn.metrics.Precision(*args, **kwargs)

Bases: ontolearn.abstracts.AbstractScorer

Precision quality function.

Attribute:

name: name of the metric = ‘Precision’.

__slots__ = ()

name: Final = 'Precision'

score2(tp: int, fn: int, fp: int, tn: int) → Tuple[bool, float]

Quality score for a coverage count.

Parameters:

• tp – True positive count.

• fn – False negative count.

• fp – False positive count.

• tn – True negative count.

Returns:

Tuple, first position indicating if the function could be applied, second position the quality value

in the range 0.0–1.0.

class ontolearn.metrics.F1(*args, **kwargs)

Bases: ontolearn.abstracts.AbstractScorer

F1-score quality function.

Attribute:

name: name of the metric = ‘F1’.

__slots__ = ()

name: Final = 'F1'

score2(tp: int, fn: int, fp: int, tn: int) → Tuple[bool, float]

Quality score for a coverage count.

Parameters:

• tp – True positive count.

• fn – False negative count.

• fp – False positive count.

• tn – True negative count.

Returns:

Tuple, first position indicating if the function could be applied, second position the quality value

in the range 0.0–1.0.

class ontolearn.metrics.Accuracy(*args, **kwargs)

Bases: ontolearn.abstracts.AbstractScorer

Accuracy quality function. Accuracy is acc = (tp + tn) / (tp + tn + fp+ fn). However, Concept learning papers (e.g. Learning OWL Class expression) appear to invent their own accuracy metrics.

In OCEL => Accuracy of a concept = 1 - ( |E^+ R(C)|+ |E^- AND R(C)|) / |E|).

In CELOE => Accuracy of a concept C = 1 - ( |R(A) R(C)| + |R(C) R(A)|)/n.

1. R(.) is the retrieval function, A is the class to describe and C in CELOE.

2. E^+ and E^- are the positive and negative examples probided. E = E^+ OR E^- .

Attribute:

name: name of the metric = ‘Accuracy’.

__slots__ = ()

name: Final = 'Accuracy'

score2(tp: int, fn: int, fp: int, tn: int) → Tuple[bool, float]

Quality score for a coverage count.

Parameters:

• tp – True positive count.

• fn – False negative count.

• fp – False positive count.

• tn – True negative count.

Returns:

Tuple, first position indicating if the function could be applied, second position the quality value

in the range 0.0–1.0.

class ontolearn.metrics.WeightedAccuracy(*args, **kwargs)

Bases: ontolearn.abstracts.AbstractScorer

WeightedAccuracy quality function.

Attribute:

name: name of the metric = ‘WeightedAccuracy’.

__slots__ = ()

name: Final = 'WeightedAccuracy'

score2(tp: int, fn: int, fp: int, tn: int) → Tuple[bool, float]

Quality score for a coverage count.

Parameters:

• tp – True positive count.

• fn – False negative count.

• fp – False positive count.

• tn – True negative count.

Returns:

Tuple, first position indicating if the function could be applied, second position the quality value

in the range 0.0–1.0.