libsent/src/ngram/ngram_access.c

単語列・クラス列の N-gram 確率を求める [詳細]

#include <sent/stddefs.h>
#include <sent/ngram2.h>

ngram_access.cのインクルード依存関係図

ソースコードを見る。

関数
NNID	search_bigram (NGRAM_INFO *ndata, WORD_ID w_l, WORD_ID w_r)
static NNID	search_trigram_v3 (NGRAM_INFO *ndata, NNID n2, WORD_ID wkey)
static NNID	search_trigram_v4 (NGRAM_INFO *ndata, NNID n2, WORD_ID wkey)
LOGPROB	uni_prob (NGRAM_INFO *ndata, WORD_ID w)
LOGPROB	bi_prob_lr (NGRAM_INFO *ndata, WORD_ID w1, WORD_ID w2)
LOGPROB	bi_prob_rl (NGRAM_INFO *ndata, WORD_ID w1, WORD_ID w2)
LOGPROB	tri_prob_rl (NGRAM_INFO *ndata, WORD_ID w1, WORD_ID w2, WORD_ID w3)

---

説明

単語列・クラス列の N-gram 確率を求める

作者:

Akinobu LEE

日付:

Wed Feb 16 07:46:18 2005

Revision

1.5

ngram_access.c で定義されています。

---

関数

NNID search_bigram	(	NGRAM_INFO *	ndata,
		WORD_ID	w_l,
		WORD_ID	w_r
	)

Search for 2-gram tuple (w_l, w_r) in the 2-gram part of N-gram.

引数:

	ndata	[in] word/class N-gram
	w_l	[in] left word/class ID in N-gram
	w_r	[in] right word/class ID in N-gram

戻り値:

corresponding index to the 2-gram data part if found, or NNID_INVALID if the tuple does not exist in 2-gram.

ngram_access.c の 39 行で定義されています。

参照元 add_bigram_rl(), bi_prob_lr(), bi_prob_rl(), set_trigram(), と tri_prob_rl().

static NNID search_trigram_v3	(	NGRAM_INFO *	ndata,
		NNID	n2,
		WORD_ID	wkey
	)			[static]

Search for a reverse 3-gram tuple (wkey, w1, w2), where the context 2-gram tuple (w1, w2) should exist on 2-gram data part at n2.

(for v3 struture format)

引数:

	ndata	[in] word/class N-gram
	n2	[in] 2-gram data part where context 2-gram data exists.
	wkey	[in] left word/class ID in N-gram

戻り値:

corresponding index to the 3-gram data part if found, or NNID_INVALID if the 3-gram does not exist.

ngram_access.c の 77 行で定義されています。

参照元 tri_prob_rl().

static NNID search_trigram_v4	(	NGRAM_INFO *	ndata,
		NNID	n2,
		WORD_ID	wkey
	)			[static]

Search for a reverse 3-gram tuple (wkey, w1, w2), where the context 2-gram tuple (w1, w2) should exist on 2-gram data part at n2.

(for v4 struture format)

引数:

	ndata	[in] word/class N-gram
	n2	[in] 2-gram data part where context 2-gram data exists.
	wkey	[in] left word/class ID in N-gram

戻り値:

corresponding index to the 3-gram data part if found, or NNID_INVALID if the 3-gram does not exist.

ngram_access.c の 115 行で定義されています。

参照元 tri_prob_rl().

LOGPROB uni_prob	(	NGRAM_INFO *	ndata,
		WORD_ID	w
	)

Get 1-gram probability of in log10.

引数:

	ndata	[in] word/class N-gram
	w	[in] word/class ID in N-gram

戻り値:

log10 probability .

ngram_access.c の 154 行で定義されています。

参照元 build_wchmm2(), get_nbest_uniprob(), max_successor_prob(), と pick_backtrellis_words().

LOGPROB bi_prob_lr	(	NGRAM_INFO *	ndata,
		WORD_ID	w1,
		WORD_ID	w2
	)

Get LR 2-gram probability of word/class sequence in log10

引数:

	ndata	[in] word/class N-gram
	w1	[in] left word/class ID in N-gram
	w2	[in] right word/class ID in N-gram (to compute probability)

戻り値:

log10 probability .

ngram_access.c の 175 行で定義されています。

参照元 calc_successor_prob(), max_successor_prob(), と max_successor_prob_iw().

LOGPROB bi_prob_rl	(	NGRAM_INFO *	ndata,
		WORD_ID	w1,
		WORD_ID	w2
	)

Get RL 2-gram probability of word/class sequence in log10.

引数:

	ndata	[in] word/class N-gram
	w1	[in] left word/class ID in N-gram (to compute probability)
	w2	[in] right word/class ID in N-gram

戻り値:

log10 probability .

ngram_access.c の 206 行で定義されています。

参照元 pick_backtrellis_words(), と tri_prob_rl().

LOGPROB tri_prob_rl	(	NGRAM_INFO *	ndata,
		WORD_ID	w1,
		WORD_ID	w2,
		WORD_ID	w3
	)

Get RL 3-gram probability of word/class sequence in log10.

引数:

	ndata	[in] word/class N-gram
	w1	[in] left word/class ID in N-gram (to compute probability)
	w2	[in] middle word/class ID in N-gram
	w3	[in] right word/class ID in N-gram

戻り値:

log10 probability .

ngram_access.c の 239 行で定義されています。

参照元 pick_backtrellis_words().