word2vec

import jieba

s1 = '这只皮靴号码大了,那只号码合适'
s2 = '这只皮靴号码不小,那只更合适'

def s_seg(s):
    s_seg = '/'.join([x for x in jieba.cut(s,cut_all=True) if x!=''])
    s_lst = [x for x in jieba.cut(s,cut_all=True) if x!='']
    s_set = set(s_lst)
    return s_seg,s_lst,s_set

s1_seg,s1_lst,s1_set = s_seg(s1)
s2_seg,s2_lst,s2_set = s_seg(s2)

#词库
s_dict = s1_set.union(s2_set)

#词库索引
i = 0
word_dict = dict()
for word in s_dict:
    word_dict[word] = i
    i += 1
print(word_dict)

def word_2_vec(word_dict,s_list):
    word_cnt = dict()
    s_vector = [0] * len(word_dict)

    #word count
    for word in s_list:
        if word_cnt.get(word, -1) == -1:
            word_cnt[word] = 1
        else:
            word_cnt[word] += 1
    # print(word_cnt)

    # vector
    for word,freq in word_cnt.items():
        word_id = word_dict[word]
        s_vector[word_id] = freq
    print(s_vector)

s1_vec = word_2_vec(word_dict, s1_lst)
s2_vec = word_2_vec(word_dict, s2_lst)