summaryrefslogtreecommitdiffstats
path: root/utils/training/estimate_k_mixture_model.cpp
blob: f5c5777ee2946cc2ec6fde56036ac9356a30c662 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
/* 
 *  libpinyin
 *  Library to deal with pinyin.
 *  
 *  Copyright (C) 2011 Peng Wu <alexepico@gmail.com>
 *  
 *  This program is free software: you can redistribute it and/or modify
 *  it under the terms of the GNU General Public License as published by
 *  the Free Software Foundation, either version 3 of the License, or
 *  (at your option) any later version.
 *
 *  This program is distributed in the hope that it will be useful,
 *  but WITHOUT ANY WARRANTY; without even the implied warranty of
 *  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 *  GNU General Public License for more details.
 *
 *  You should have received a copy of the GNU General Public License
 *  along with this program.  If not, see <http://www.gnu.org/licenses/>.
 */

#include <locale.h>
#include "pinyin_internal.h"
#include "k_mixture_model.h"

static const gchar * bigram_filename = "k_mixture_model_ngram.db";
static const gchar * deleted_bigram_filename = "k_mixture_model_deleted_ngram.db";

static GOptionEntry entries[] =
{
    {"bigram-file", 0, 0, G_OPTION_ARG_FILENAME, &bigram_filename, "the bigram file", NULL},
    {"deleted-bigram-file", 0, 0, G_OPTION_ARG_FILENAME, &deleted_bigram_filename, "the deleted bigram file", NULL},
    {NULL}
};


parameter_t compute_interpolation(KMixtureModelSingleGram * deleted_bigram,
                                  KMixtureModelBigram * unigram,
                                  KMixtureModelSingleGram * bigram){
    bool success;
    parameter_t lambda = 0, next_lambda = 0.6;
    parameter_t epsilon = 0.001;

    KMixtureModelMagicHeader magic_header;
    check_result(unigram->get_magic_header(magic_header));
    assert(0 != magic_header.m_total_freq);

    while (fabs(lambda - next_lambda) > epsilon){
        lambda = next_lambda;
        next_lambda = 0;
        parameter_t numerator = 0;
        parameter_t part_of_denominator = 0;

        FlexibleBigramPhraseArray array = g_array_new(FALSE, FALSE, sizeof(KMixtureModelArrayItemWithToken));
        deleted_bigram->retrieve_all(array);

        for ( size_t i = 0; i < array->len; ++i){
            KMixtureModelArrayItemWithToken * item = &g_array_index(array, KMixtureModelArrayItemWithToken, i);
            //get the phrase token
            phrase_token_t token = item->m_token;
            guint32 deleted_count = item->m_item.m_WC;

            {
                parameter_t elem_poss = 0;
                KMixtureModelArrayHeader array_header;
                KMixtureModelArrayItem array_item;
                if ( bigram && bigram->get_array_item(token, array_item) ){
                    check_result(bigram->get_array_header(array_header));
                    assert(0 != array_header.m_WC);
                    elem_poss = array_item.m_WC / (parameter_t) array_header.m_WC;
                }
                numerator = lambda * elem_poss;
            }

            {
                parameter_t elem_poss = 0;
                KMixtureModelArrayHeader array_header;
                if (unigram->get_array_header(token, array_header)){
                    elem_poss = array_header.m_freq / (parameter_t) magic_header.m_total_freq;
                }
                part_of_denominator = (1 - lambda) * elem_poss;
            }
            if (0 == (numerator + part_of_denominator))
                continue;

            next_lambda += deleted_count * (numerator / (numerator + part_of_denominator));
        }
        KMixtureModelArrayHeader header;
        check_result(deleted_bigram->get_array_header(header));
        assert(0 != header.m_WC);
        next_lambda /= header.m_WC;

        g_array_free(array, TRUE);
    }
    lambda = next_lambda;
    return lambda;
}

int main(int argc, char * argv[]){
    setlocale(LC_ALL, "");

    GError * error = NULL;
    GOptionContext * context;

    context = g_option_context_new("- estimate k mixture model");
    g_option_context_add_main_entries(context, entries, NULL);
    if (!g_option_context_parse(context, &argc, &argv, &error)) {
        g_print("option parsing failed:%s\n", error->message);
        exit(EINVAL);
    }

    /* magic header signature check here. */
    KMixtureModelBigram bigram(K_MIXTURE_MODEL_MAGIC_NUMBER);
    bigram.attach(bigram_filename, ATTACH_READONLY);

    KMixtureModelBigram deleted_bigram(K_MIXTURE_MODEL_MAGIC_NUMBER);
    deleted_bigram.attach(deleted_bigram_filename, ATTACH_READONLY);

    GArray * deleted_items = g_array_new(FALSE, FALSE, sizeof(phrase_token_t));
    deleted_bigram.get_all_items(deleted_items);

    parameter_t lambda_sum = 0;
    int lambda_count = 0;

    for( size_t i = 0; i < deleted_items->len; ++i ){
        phrase_token_t * token = &g_array_index(deleted_items, phrase_token_t, i);
        KMixtureModelSingleGram * single_gram = NULL;
        bigram.load(*token, single_gram, true);

        KMixtureModelSingleGram * deleted_single_gram = NULL;
        deleted_bigram.load(*token, deleted_single_gram);

        KMixtureModelArrayHeader array_header;
        if (single_gram)
            check_result(single_gram->get_array_header(array_header));
        KMixtureModelArrayHeader deleted_array_header;
        check_result(deleted_single_gram->get_array_header(deleted_array_header));

        if ( 0 != deleted_array_header.m_WC ) {
            parameter_t lambda = compute_interpolation(deleted_single_gram, &bigram, single_gram);

            printf("token:%d lambda:%f\n", *token, lambda);

            lambda_sum += lambda;
            lambda_count ++;
        }

        if (single_gram)
            delete single_gram;
        delete deleted_single_gram;
    }

    printf("average lambda:%f\n", (lambda_sum/lambda_count));
    g_array_free(deleted_items, TRUE);
    return 0;
}