python 字符相似度对比

1，字

　　将字符以逗号，句号为一个段落，在对段落内的字符同一字进行对比

2，词

　　同一词语进行对比

 

import difflib
def stri_similar(s1,s2):
    return difflib.SequenceMatcher(None,s1,s2).quick_ratio()
data1 = '你好啊'
data2 = '你好'
# for i in range(len(data1)):
#     s1 = data1[i]
#     s2 = data2[i]
#     print(stri_similar(s1,s2) )
# print(stri_similar(data1,data2) )

#相似比对比
#顺序对比法：两个列表按顺序对比 缺点：计算量小 ； 优点：相似度对比准确度低
#遍历对比法：一个列表遍历另一个列表的全部，一段对比每一段。 缺点：计算量大 ； 优点：相似度对比更有准确度
d = ['长短搭配，', '尽量减少接头，', '以节约钢材。']
text = ['搭配，', '尽少接头，', '以节约钢材。']

def contrast(text,d):
    try:
        count = 0
        for i in range(len(text)): #遍历段落
            count1 = 0
            if len(text[i]) > len(d[i]):
                c = text[i]
                t = d[i]
            else :
                c = d[i]
                t = text[i]

            for p in c: #遍历段落字符
                if p in t:
                    count1 += 1
            count += count1/len(c)
        # print('{:.2f} {:.2%}'.format(count,count/len(text)))
        return count/len(text)
    except:
        print('报错：列表长度不一样')
# contrast(text,d)

def contrast2(list1,list2):
    count = 0
    for i in list1:
        count1 = 0
        for t in list2:
            count1 += contrast([t],[i])
        count += count1
    print('总量：{:.2f} 相似度：{:.2%}'.format(count,count/len(list1)))
list1 = ['长短搭配，', '尽量减少接头，', '以节约钢材。']
list2 = ['长短搭配，','尽量减少接头，']
contrast2(list1,list2)
# contrast(list1,list2)
word_list = []
s_vector = []
import jieba
s1 = '我喜欢你啊'
s2 = '我喜欢你'
def max_len(s1,s2):
    a,b = len(s1),len(s2)
    if a > b:
        data =s1
    else:
        data = s2
    return data
print(max_len(s1,s2) )
#第一种补零
#第二种中和
for i in list(jieba.cut(s1)):
    print(i)
def merge(s1,s2):
    word_list = list(jieba.cut(s1)) #词典
    for i,d in list(jieba.cut(s2)):
        if i in word_list :
            continue
        else:
            word_list.append(i)
# 第一步
def vector(data): #词典匹配 word_list
    for i in word_list:
        if i in list(jieba.cut(data)):
            s_vector.append(1)
        else:
            s_vector.append(0)
# 第二步
def cosine_Similarity(s1,s2): #余弦
    l,l1,l2 = 0,0,0
    for i,d in enumerate(s1):
        l += s1[i]*s2[i]
        l1 += s1[i]**2
        l2 += s2[i]**2
    return  l/((l1**0.5)*(l2**0.5))
def Euclidean_distance(s1,s2): #欧氏距离
    vector_list = 0
    for i,d in enumerate(s1):
        vector_list += (s1[i]-s2[i])**2
    return vector_list**0.5

print((2*2)**0.5,2**2)
s1 = [1,1,1,1,1]
s2 = [3,3,3,3,3]
d = Euclidean_distance(s1,s2)
print(d)
d = cosine_Similarity(s1,s2)
print(d)