python提取内容关键词的方法

本文实例讲述了python提取内容关键词的方法。分享给大家供大家参考。具体分析如下:

一个非常高效的提取内容关键词的python代码,这段代码只能用于英文文章内容,中文因为要分词,这段代码就无能为力了,不过要加上分词功能,效果和英文是一样的。


# coding=UTF-8

import nltk

from nltk.corpus import brown

# This is a fast and simple noun phrase extractor (based on NLTK)

# Feel free to use it, just keep a link back to this post

# http://thetokenizer.com/2013/05/09/efficient-way-to-extract-the-main-topics-of-a-sentence/

# Create by Shlomi Babluki

# May, 2013

  

# This is our fast Part of Speech tagger

#############################################################################

brown_train = brown.tagged_sents(categories='news')

regexp_tagger = nltk.RegexpTagger(

    [(r'^-?[0-9]+(.[0-9]+)?$', 'CD'),

     (r'(-|:|;)$', ':'),

     (r'\'*$', 'MD'),

     (r'(The|the|A|a|An|an)$', 'AT'),

     (r'.*able$', 'JJ'),

     (r'^[A-Z].*$', 'NNP'),

     (r'.*ness$', 'NN'),

     (r'.*ly$', 'RB'),

     (r'.*s$', 'NNS'),

     (r'.*ing$', 'VBG'),

     (r'.*ed$', 'VBD'),

     (r'.*', 'NN')

])

unigram_tagger = nltk.UnigramTagger(brown_train, backoff=regexp_tagger)

bigram_tagger = nltk.BigramTagger(brown_train, backoff=unigram_tagger)

#############################################################################

# This is our semi-CFG; Extend it according to your own needs

#############################################################################

cfg = {}

cfg["NNP+NNP"] = "NNP"

cfg["NN+NN"] = "NNI"

cfg["NNI+NN"] = "NNI"

cfg["JJ+JJ"] = "JJ"

cfg["JJ+NN"] = "NNI"

#############################################################################

class NPExtractor(object):

    def __init__(self, sentence):

        self.sentence = sentence

    # Split the sentence into singlw words/tokens

    def tokenize_sentence(self, sentence):

        tokens = nltk.word_tokenize(sentence)

        return tokens

    # Normalize brown corpus' tags ("NN", "NN-PL", "NNS" > "NN")

    def normalize_tags(self, tagged):

        n_tagged = []

        for t in tagged:

            if t[1] == "NP-TL" or t[1] == "NP":

                n_tagged.append((t[0], "NNP"))

                continue

            if t[1].endswith("-TL"):

                n_tagged.append((t[0], t[1][:-3]))

                continue

            if t[1].endswith("S"):

                n_tagged.append((t[0], t[1][:-1]))

                continue

            n_tagged.append((t[0], t[1]))

        return n_tagged

    # Extract the main topics from the sentence

    def extract(self):

        tokens = self.tokenize_sentence(self.sentence)

        tags = self.normalize_tags(bigram_tagger.tag(tokens))

        merge = True

        while merge:

            merge = False

            for x in range(0, len(tags) - 1):

                t1 = tags[x]

                t2 = tags[x + 1]

                key = "%s+%s" % (t1[1], t2[1])

                value = cfg.get(key, '')

                if value:

                    merge = True

                    tags.pop(x)

                    tags.pop(x)

                    match = "%s %s" % (t1[0], t2[0])

                    pos = value

                    tags.insert(x, (match, pos))

                    break

        matches = []

        for t in tags:

            if t[1] == "NNP" or t[1] == "NNI":

            #if t[1] == "NNP" or t[1] == "NNI" or t[1] == "NN":

                matches.append(t[0])

        return matches

# Main method, just run "python np_extractor.py"

def main():

    sentence = "Swayy is a beautiful new dashboard for discovering and curating online content."

    np_extractor = NPExtractor(sentence)

    result = np_extractor.extract()

    print "This sentence is about: %s" % ", ".join(result)

if __name__ == '__main__':

    main()

希望本文所述对大家的Python程序设计有所帮助。

声明:本文内容来源于网络,版权归原作者所有,内容由互联网用户自发贡献自行上传,本网站不拥有所有权,未作人工编辑处理,也不承担相关法律责任。如果您发现有涉嫌版权的内容,欢迎发送邮件至:notice#cainiaojc.com(发邮件时,请将#更换为@)进行举报,并提供相关证据,一经查实,本站将立刻删除涉嫌侵权内容。