idioma Class Reference

Class "idioma" implements a visible Markov's model that calculates the probability that a text is in a certain language. More...

#include <idioma.h>

List of all members.

Public Member Functions
	idioma ()
	null constructor
	idioma (const std::wstring &)
	Constructor, given the model file to load.
double	sequence_probability (std::wistream &, size_t &) const
	Calculates the probability that the text is in the instance language.
double	compute_probability (const std::wstring &, double s=1.0) const
	Compute normalized language probability for given string.
void	train (const std::wstring &, const std::wstring &, const std::wstring &)
	Create a new model for the language from given input file, Store model in given filename, with given language code.
void	train (std::wistream &f, const std::wstring &, const std::wstring &)
std::wstring	get_language_code () const
	get iso code for current language
Private Member Functions
std::wstring	from_writable (const std::wstring &) const
	convert a trigram from writable represntation in the model file
std::wstring	to_writable (const std::wstring &) const
	convert a trigram to a writable represntation for the model file
double	ProbA (const std::wstring &, const std::wstring &) const
	Consult method for transition probabilities.
double	ProbPi (const std::wstring &) const
	Consult method for initial probabilities.
void	increment (std::map< std::wstring, double > &, const std::wstring &)
	Increase occurrences of a n-gram.
void	increment (std::map< std::pair< std::wstring, std::wstring >, double > &, const std::wstring &, const std::wstring &)
	Increase occurrences of a two chained trigrams.
void	initial_trigram (std::wistream &, wchar_t &, wchar_t &, wchar_t &) const
	Initial trigram: two fictitious ' ' plus the first actual letter.
std::wstring	trigram (wchar_t, wchar_t, wchar_t) const
	build actual trigram from iterators
void	create_model (std::wistream &f)
	Create new model from given stream, with given language code.
void	save_model (const std::wstring &) const
	Save current model in given file.
Private Attributes
std::wstring	LangCode
std::map< std::wstring, double >	pa_nom
	State transitions probabilities.
std::map< std::wstring, double >	ppi_nom
	Initial probabilities.
std::map< std::wstring, double >	pi
	auxiliary for training
std::map< std::wstring, double >	A
std::map< std::pair < std::wstring, std::wstring > , double >	B
	auxiliary for training
size_t	nf
	auxiliary for training

---

Detailed Description

Class "idioma" implements a visible Markov's model that calculates the probability that a text is in a certain language.

---

Constructor & Destructor Documentation

idioma::idioma

(

)

null constructor

idioma::idioma

(

const std::wstring &

)

Constructor, given the model file to load.

---

Member Function Documentation

double idioma::compute_probability	(	const std::wstring &	,
		double	s = 1.0
	)		const

Compute normalized language probability for given string.

void idioma::create_model

(

std::wistream &

f

)

[private]

Create new model from given stream, with given language code.

std::wstring idioma::from_writable

(

const std::wstring &

)

const [private]

convert a trigram from writable represntation in the model file

std::wstring idioma::get_language_code

(

)

const

get iso code for current language

void idioma::increment	(	std::map< std::wstring, double > &	,
		const std::wstring &
	)		[private]

Increase occurrences of a n-gram.

void idioma::increment	(	std::map< std::pair< std::wstring, std::wstring >, double > &	,
		const std::wstring &	,
		const std::wstring &
	)		[private]

Increase occurrences of a two chained trigrams.

void idioma::initial_trigram	(	std::wistream &	,
		wchar_t &	,
		wchar_t &	,
		wchar_t &
	)		const [private]

Initial trigram: two fictitious '
' plus the first actual letter.

double idioma::ProbA	(	const std::wstring &	,
		const std::wstring &
	)		const [private]

Consult method for transition probabilities.

double idioma::ProbPi

(

const std::wstring &

)

const [private]

Consult method for initial probabilities.

void idioma::save_model

(

const std::wstring &

)

const [private]

Save current model in given file.

double idioma::sequence_probability	(	std::wistream &	,
		size_t &
	)		const

Calculates the probability that the text is in the instance language.

std::wstring idioma::to_writable

(

const std::wstring &

)

const [private]

convert a trigram to a writable represntation for the model file

void idioma::train	(	const std::wstring &	,
		const std::wstring &	,
		const std::wstring &
	)

Create a new model for the language from given input file, Store model in given filename, with given language code.

void idioma::train	(	std::wistream &	f,
		const std::wstring &	,
		const std::wstring &
	)

std::wstring idioma::trigram	(	wchar_t	,
		wchar_t	,
		wchar_t
	)		const [private]

build actual trigram from iterators

---

Member Data Documentation

std::map<std::wstring,double> idioma::A [private]

std::map<std::pair<std::wstring,std::wstring>,double> idioma::B [private]

auxiliary for training

std::wstring idioma::LangCode [private]

size_t idioma::nf [private]

auxiliary for training

std::map<std::wstring,double> idioma::pa_nom [private]

State transitions probabilities.

std::map<std::wstring,double> idioma::pi [private]

auxiliary for training

std::map<std::wstring,double> idioma::ppi_nom [private]

Initial probabilities.

---

The documentation for this class was generated from the following file:

• idioma.h