python实现textrank代码

好的，下面是 Python 实现的 TextRank 代码：

import itertools
import networkx as nx
from nltk.tokenize import sent_tokenize, word_tokenize
from nltk.corpus import stopwords
from string import punctuation

def preprocess(text):
    stop_words = set(stopwords.words('english') + list(punctuation))
    words = [word.lower() for word in word_tokenize(text)]
    return [word for word in words if word not in stop_words]

def similarity(sent1, sent2):
    words1 = set(sent1)
    words2 = set(sent2)
    return len(words1 & words2) / (len(words1) + len(words2))

def textrank(text, top_n):
    sentences = sent_tokenize(text)
    sentence_tokens = [preprocess(sent) for sent in sentences]
    sim_graph = nx.Graph()
    for sent1, sent2 in itertools.product(sentence_tokens, repeat=2):
        if sent1 != sent2:
            sim_graph.add_edge(sentences.index(' '.join(sent1)), sentences.index(' '.join(sent2)), weight=similarity(sent1, sent2))
    scores = nx.pagerank(sim_graph)
    sorted_scores = sorted(scores.items(), key=lambda x: x[1], reverse=True)
    return [sentences[idx] for idx, score in sorted_scores[:top_n]]

以上代码实现了基于 TextRank 算法的文本摘要功能，它将文本分成若干句子，通过计算句子之间的相似度来建立句子之间的图，然后对图进行 PageRank 计算，得到每个句子的重要性分数，并按照分数的降序返回前 top_n 个句子作为摘要。