python实现过滤敏感词

时间:2022-09-07 00:27:40

简述:

关于敏感词过滤可以看成是一种文本反垃圾算法,例如
 题目:敏感词文本文件 filtered_words.txt,当用户输入敏感词语,则用 星号 * 替换,例如当用户输入「北京是个好城市」,则变成「**是个好城市」
 代码:

?
1
2
3
4
5
6
7
8
9
10
11
12
13
#coding=utf-8
def filterwords(x):
    with open(x,'r') as f:
        text=f.read()
    print text.split('\n')
    userinput=raw_input('myinput:')
    for i in text.split('\n'):
        if i in userinput:
            replace_str='*'*len(i.decode('utf-8'))
            word=userinput.replace(i,replace_str)
            return word
 
print filterwords('filtered_words.txt')

再例如反黄系列:

?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
开发敏感词语过滤程序,提示用户输入评论内容,如果用户输入的内容中包含特殊的字符:
敏感词列表 li = ["苍老师","东京热",”武藤兰”,”波多野结衣”]
则将用户输入的内容中的敏感词汇替换成***,并添加到一个列表中;如果用户输入的内容没有敏感词汇,则直接添加到上述的列表中。
content = input('请输入你的内容:')
li = ["苍老师","东京热","武藤兰","波多野结衣"]
i = 0
while i < 4:
    for li[i] in content:
        li1 = content.replace('苍老师','***')
        li2 = li1.replace('东京热','***')
        li3 = li2.replace('武藤兰','***')
        li4 = li3.replace('波多野结衣','***')
    else:
        pass
    i += 1

python实现过滤敏感词

实战案例:

 一道bat面试题:快速替换10亿条标题中的5万个敏感词,有哪些解决思路?
 有十亿个标题,存在一个文件中,一行一个标题。有5万个敏感词,存在另一个文件。写一个程序过滤掉所有标题中的所有敏感词,保存到另一个文件中。

1、DFA过滤敏感词算法

在实现文字过滤的算法中,DFA是比较好的实现算法。DFA即Deterministic Finite Automaton,也就是确定有穷自动机。
 算法核心是建立了以敏感词为基础的许多敏感词树。
 python 实现DFA算法:

?
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
# -*- coding:utf-8 -*-
 
import time
time1=time.time()
 
# DFA算法
class DFAFilter():
    def __init__(self):
        self.keyword_chains = {}
        self.delimit = '\x00'
 
    def add(self, keyword):
        keyword = keyword.lower()
        chars = keyword.strip()
        if not chars:
            return
        level = self.keyword_chains
        for i in range(len(chars)):
            if chars[i] in level:
                level = level[chars[i]]
            else:
                if not isinstance(level, dict):
                    break
                for j in range(i, len(chars)):
                    level[chars[j]] = {}
                    last_level, last_char = level, chars[j]
                    level = level[chars[j]]
                last_level[last_char] = {self.delimit: 0}
                break
        if i == len(chars) - 1:
            level[self.delimit] = 0
 
    def parse(self, path):
        with open(path,encoding='utf-8') as f:
            for keyword in f:
                self.add(str(keyword).strip())
 
    def filter(self, message, repl="*"):
        message = message.lower()
        ret = []
        start = 0
        while start < len(message):
            level = self.keyword_chains
            step_ins = 0
            for char in message[start:]:
                if char in level:
                    step_ins += 1
                    if self.delimit not in level[char]:
                        level = level[char]
                    else:
                        ret.append(repl * step_ins)
                        start += step_ins - 1
                        break
                else:
                    ret.append(message[start])
                    break
            else:
                ret.append(message[start])
            start += 1
 
        return ''.join(ret)
 
 
if __name__ == "__main__":
    gfw = DFAFilter()
    path="F:/文本反垃圾算法/sensitive_words.txt"
    gfw.parse(path)
    text="**苹果新品发布会雞八"
    result = gfw.filter(text)
 
    print(text)
    print(result)
    time2 = time.time()
    print('总共耗时:' + str(time2 - time1) + 's')

运行效果:

?
1
2
3
**苹果新品发布会雞八
****苹果新品发布会**
总共耗时:0.0010344982147216797s

2、AC自动机过滤敏感词算法

AC自动机:一个常见的例子就是给出n个单词,再给出一段包含m个字符的文章,让你找出有多少个单词在文章里出现过。
 简单地讲,AC自动机就是字典树+kmp算法+失配指针

?
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
# -*- coding:utf-8 -*-
 
import time
time1=time.time()
 
# AC自动机算法
class node(object):
    def __init__(self):
        self.next = {}
        self.fail = None
        self.isWord = False
        self.word = ""
 
class ac_automation(object):
 
    def __init__(self):
        self.root = node()
 
    # 添加敏感词函数
    def addword(self, word):
        temp_root = self.root
        for char in word:
            if char not in temp_root.next:
                temp_root.next[char] = node()
            temp_root = temp_root.next[char]
        temp_root.isWord = True
        temp_root.word = word
 
    # 失败指针函数
    def make_fail(self):
        temp_que = []
        temp_que.append(self.root)
        while len(temp_que) != 0:
            temp = temp_que.pop(0)
            p = None
            for key,value in temp.next.item():
                if temp == self.root:
                    temp.next[key].fail = self.root
                else:
                    p = temp.fail
                    while p is not None:
                        if key in p.next:
                            temp.next[key].fail = p.fail
                            break
                        p = p.fail
                    if p is None:
                        temp.next[key].fail = self.root
                temp_que.append(temp.next[key])
 
    # 查找敏感词函数
    def search(self, content):
        p = self.root
        result = []
        currentposition = 0
 
        while currentposition < len(content):
            word = content[currentposition]
            while word in p.next == False and p != self.root:
                p = p.fail
 
            if word in p.next:
                p = p.next[word]
            else:
                p = self.root
 
            if p.isWord:
                result.append(p.word)
                p = self.root
            currentposition += 1
        return result
 
    # 加载敏感词库函数
    def parse(self, path):
        with open(path,encoding='utf-8') as f:
            for keyword in f:
                self.addword(str(keyword).strip())
 
    # 敏感词替换函数
    def words_replace(self, text):
        """
        :param ah: AC自动机
        :param text: 文本
        :return: 过滤敏感词之后的文本
        """
        result = list(set(self.search(text)))
        for x in result:
            m = text.replace(x, '*' * len(x))
            text = m
        return text
 
 
 
 
 
if __name__ == '__main__':
 
    ah = ac_automation()
    path='F:/文本反垃圾算法/sensitive_words.txt'
    ah.parse(path)
    text1="**苹果新品发布会雞八"
    text2=ah.words_replace(text1)
 
    print(text1)
    print(text2)
 
    time2 = time.time()
    print('总共耗时:' + str(time2 - time1) + 's')

运行结果:

?
1
2
3
**苹果新品发布会雞八
****苹果新品发布会**
总共耗时:0.0010304450988769531s

以上就是python实现过滤敏感词的详细内容,更多关于python 过滤敏感词的资料请关注服务器之家其它相关文章!

原文链接:https://cloud.tencent.com/developer/article/1395616