Package nltk :: Module align :: Class AlignedSent

type AlignedSent

object --+
         |
        AlignedSent

Aligned sentence object. Encapsulates two sentences along with an Alignment between them.

Instance Methods

[hide private]

__init__(self, words=[], mots=[], alignment='', encoding='latin-1')
Initialize a new AlignedSent. source code

alignment(self, alignment)

source code

boolean

_check_align(self, a)
Returns: True if passed alignment check

source code

string

__repr__(self)
Returns: A string representation for this AlignedSent. source code

string

__str__(self)
Returns: A string representation for this AlignedSent. source code

AlignedSent

invert(self)
Returns: the invert object

source code

float or None

precision(self, reference)
Calculates the precision of an aligned sentence with respect to a "gold standard" reference AlignedSent. source code

float or None

recall(self, reference)
Calculates the recall of an aligned sentence with respect to a "gold standard" reference AlignedSent. source code

float or None

alignment_error_rate(self, reference, possible=None)
Calculates the Alignment Error Rate (AER) of an aligned sentence with respect to a "gold standard" reference AlignedSent. source code

Properties

[hide private]

words

mots

Method Details

[hide private]

init(self, words=`[]`, mots=`[]`, alignment=`''`, encoding=`'latin-1'`)
(Constructor)

source code

Initialize a new AlignedSent.

Parameters:

words (list of str) - source language words
mots (list of str) - target language words
alignment (Alignment) - the word-level alignments between the source and target language

Overrides: object.__init__

_check_align(self, a)

source code

Parameters:

a - alignment to be checked

Returns: boolean

True if passed alignment check

Raises:

IndexError - if alignment is out of sentence boundary

repr(self)
(Representation operator)

source code

Returns: string: A string representation for this AlignedSent.
Overrides: object.__repr__

str(self)
(Informal representation operator)

source code

Returns: string: A string representation for this AlignedSent.
Overrides: object.__str__

invert(self)

source code

Returns: AlignedSent: the invert object

precision(self, reference)

source code

Calculates the precision of an aligned sentence with respect to a "gold standard" reference AlignedSent.

The "possible" precision is used since it doesn't penalise for finding an alignment that was marked as "possible".

Parameters:

reference (AlignedSent or Alignment) - A "gold standard" reference aligned sentence.

Returns: float or None

recall(self, reference)

source code

Calculates the recall of an aligned sentence with respect to a "gold standard" reference AlignedSent.

The "sure" recall is used so we don't penalise for missing an alignment that was only marked as "possible".

Parameters:

reference (AlignedSent or Alignment) - A "gold standard" reference aligned sentence.

Returns: float or None

alignment_error_rate(self, reference, possible=None)

source code

Calculates the Alignment Error Rate (AER) of an aligned sentence with respect to a "gold standard" reference AlignedSent.

Return an error rate between 0.0 (perfect alignment) and 1.0 (no alignment).

Parameters:

reference (AlignedSent or Alignment) - A "gold standard" reference aligned sentence.
possible (AlignedSent or Alignment or None) - A "gold standard" reference of possible alignments (defaults to reference if None)

Returns: float or None

Property Details

[hide private]

type AlignedSent

__init__(self, words=[], mots=[], alignment='', encoding='latin-1') (Constructor)

_check_align(self, a)

__repr__(self) (Representation operator)

__str__(self) (Informal representation operator)

invert(self)

precision(self, reference)

recall(self, reference)

alignment_error_rate(self, reference, possible=None)

words

mots

init(self, words=`[]`, mots=`[]`, alignment=`''`, encoding=`'latin-1'`)
(Constructor)

repr(self)
(Representation operator)

str(self)
(Informal representation operator)