在word2vec中,有一个默认程序distance,可以用来计算给定词的最相近的top 40个词,但是不能计算给定两个词的相似程度。本程序对distance.c进行了修改,可以计算给定两个词的相似度。程序如下:
// Copyright 2013 Google Inc. All Rights Reserved.
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
#include <stdio.h>
#include <string.h>
#include <math.h>
#include <malloc.h>
const long long max_size = 2000; // max length of strings
const long long N = 10; // number of closest words that will be shown
const long long max_w = 50; // max length of vocabulary entries
int main(int argc, char **argv) {
FILE *f;
char st1[max_size];
char *bestw[N];
char file_name[max_size], st[100][max_size];
float len;
long long words, size, a, b, c, cn, bi[100];
float *M;
char *vocab;
if (argc < 2) {
printf("Usage: ./distance <FILE>\nwhere FILE contains word projections in the BINARY FORMAT\n");
return 0;
}
strcpy(file_name, argv[1]);
f = fopen(file_name, "rb");
if (f == NULL) {
printf("Input file not found\n");
return -1;
}
fscanf(f, "%lld", &words);
fscanf(f, "%lld", &size);
vocab = (char *)malloc((long long)words * max_w * sizeof(char));
for (a = 0; a < N; a++) bestw[a] = (char *)malloc(max_size * sizeof(char));
M = (float *)malloc((long long)words * (long long)size * sizeof(float));
if (M == NULL) {
printf("Cannot allocate memory: %lld MB %lld %lld\n", (long long)words * size * sizeof(float) / 1048576, words, size);
return -1;
}
for (b = 0; b < words; b++) {
a = 0;
while (1) {
vocab[b * max_w + a] = fgetc(f);
if (feof(f) || (vocab[b * max_w + a] == ' ')) break;
if ((a < max_w) && (vocab[b * max_w + a] != '\n')) a++;
}
vocab[b * max_w + a] = 0;
for (a = 0; a < size; a++) fread(&M[a + b * size], sizeof(float), 1, f);
len = 0;
for (a = 0; a < size; a++) len += M[a + b * size] * M[a + b * size];
len = sqrt(len);
for (a = 0; a < size; a++) M[a + b * size] /= len;
}
fclose(f);
while (1) {
printf("Enter two word: ");
a = 0;
while (1) {
st1[a] = fgetc(stdin);
if ((st1[a] == '\n') || (a >= max_size - 1)) {
st1[a] = 0;
break;
}
a++;
}
if (!strcmp(st1, "EXIT")) break;
cn = 0;
b = 0;
c = 0;
while (1) {
st[cn][b] = st1[c];
b++;
c++;
st[cn][b] = 0;
if (st1[c] == 0) break;
if (st1[c] == ' ') {
cn++;
b = 0;
c++;
}
}
cn++;
if(cn != 2) {
printf("please give two words\n");
continue;
}
for (a = 0; a < cn; a++) {
for (b = 0; b < words; b++) if (!strcmp(&vocab[b * max_w], st[a])) break;
if (b == words) b = -1;
bi[a] = b;
printf("Word: %s Position in vocabulary: %lld\n", st[a], bi[a]);
if (b == -1) {
printf("Out of dictionary word!\n");
break;
}
}
if (b == -1) continue;
float first[max_size], second[max_size];
for (a = 0; a < size; ++a) {
first[a] = M[a + bi[0] * size];
second[a] = M[a + bi[1] * size];
}
float product1 = 0.0, product2 = 0.0, sum12 = 0.0;
for (a = 0; a < size; ++a) {
product1 = product1 + first[a] * first[a];
product2 = product2 + second[a] * second[a];
sum12 = sum12 + first[a] * second[a];
}
float len1 = sqrt(product1);
float len2 = sqrt(product2);
float similarity = sum12/(len1 * len2);
printf("the similarity between these two words is %f\n\n", similarity);
}
return 0;
}