介绍 import numpy as np
import jieba
import jieba.posseg as pseg
,
class TextRank(对象):
,,
def 才能;__init__(自我,,句子,,窗口,α,iternum):,
,,,self.sentence =, sentence
,,,self.window =, window
,,,self.alpha =, alpha
,,,self.edge_dict =,{}, #记录节点的边连接字典,
,,,self.iternum =, iternum #迭代次数,
,
#才能对句子进行分词,
def 才能cutSentence(自我):,
,,,jieba.load_userdict (& # 39; user_dict.txt& # 39;),
,,,tag_filter =,(& # 39;一个# 39;& # 39;d # 39;, & # 39; n # 39;, & # 39; v # 39;],
,,,seg_result =, pseg.cut (self.sentence),
,,,self.word_list =, (s.word for s 拷贝seg_result if s.flag 拷贝tag_filter),
,,,印刷(self.word_list),
,
#才能根据窗口,构建每个节点的相邻节点,返回边的集合,
def 才能createNodes(自我):,
,,,tmp_list =, [],
,,,word_list_len =, len (self.word_list),
,,,for 指数,word 列举(self.word_list):拷贝,
,,,,,if word not 拷贝self.edge_dict.keys ():,
,,,,,,,tmp_list.append(词),
,,,,,,,tmp_set =,设置(),
,,,,,,,left =, index 作用;self.window +, 1 #窗口左边界,
,,,,,,,right =, index +, self.window #窗口右边界,
,,,,,,,if left & lt;, 0:, left =, 0,
,,,,,,,if right 祝辞=,word_list_len:, right =, word_list_len
,,,,,,,for 小姐:拷贝范围(左,右):,
,,,,,,,,,if 小姐:==,指数:,
,,,,,,,,,,,continue
,,,,,,,,,tmp_set.add (self.word_list[我]),
,,,,,,,self.edge_dict[词],=,tmp_set
,
#才能根据边的相连关系,构建矩阵,
def 才能createMatrix(自我):,
,,,self.matrix =, np.zeros ((len(设置(self.word_list)), len(设置(self.word_list)))),
,,,self.word_index =,{} #记录词的index
,,,self.index_dict =,{} #记录节点指数对应的词,
,
,,,for 我,v 列举(设置(self.word_list)):拷贝,
,,,,,self.word_index [v],=,小姐:
,,,,,self.index_dict[我],=,v
,,,for key 拷贝self.edge_dict.keys ():,
,,,,,for w 拷贝self.edge_dict(例子):,
,,,,,,,self.matrix [self.word_index[主要]][self.word_index [w]],=, 1,
,,,,,,,self.matrix [self.word_index [w]] [self.word_index[主要]],=,1,
,,,#归一化,
,,,for j 拷贝范围(self.matrix.shape [1]):,
,,,,,sum =, 0,
,,,,,for 小姐:拷贝范围(self.matrix.shape [0]):,
,,,,,,,sum +=, self.matrix[我][j],
,,,,,for 小姐:拷贝范围(self.matrix.shape [0]):,
,,,,,,,self.matrix[我][j],/=, sum
,
#才能根据textrank公式计算权重,
def 才能calPR(自我):,
,,,self.PR =, np.ones ((len(设置(self.word_list)), 1]),
,,,for 小姐:范围(self.iternum):拷贝,
,,,,,self.PR =,(1,安康;self.alpha), +, self.alpha *, np.dot (self.matrix, self.PR),
,
#输才能出词和相应的权重,
def 才能printResult(自我):,
,,,word_pr =, {},
,,,for 小姐:范围(len (self.PR)):拷贝,
null
null
null
null
null
null
null
null
null
null
null
null
这篇文章主要介绍python如何实现textrank关键词提取,文中介绍的非常详细,具有一定的参考价值,感兴趣的小伙伴们一定要看完!
用python写了一个简单版本的textrank,实现提取关键词的功能。
