Text keyword extraction: ansj

background

Because in the text content , You need to extract keywords and give them to seo As a key word .

At the same time, the content is in Chinese , Need to find a corresponding Chinese word segmentation and extraction tool .

ansj

According to the official introduction , I think it's quite powerful .

Official statement ：

 This is a base n-Gram+CRF+HMM Chinese word segmentation of java Realization .

 The speed of word segmentation is about 200 Around the word （mac air Next test ）, The accuracy can reach 96% above .

 At present, Chinese word segmentation has been realized 、 Chinese name recognition 、 User defined Dictionary 、 Keyword extraction 、 Automatic summarization 、 Keyword tags and other functions .

 It can be applied to natural language processing and so on , It is suitable for all kinds of projects that require high segmentation effect .

Try it out

maven

<dependency>
    <groupId>org.ansj</groupId>
    <artifactId>ansj_seg</artifactId>
    <version>5.1.1</version>
</dependency>

Then pass in the text to call . The calling code has only the following two lines

// Set the number of keywords 
KeyWordComputer kwc = new KeyWordComputer(5);
 // Text extraction keyword .title It's the title ;content Text 
 Collection<Keyword> result = kwc.computeArticleTfidf(title, content);

Extract examples

/**
	 *  Examples of keyword extraction 
	 * @author ansj
	 *
	 */
	public class KeyWordCompuerDemo {
	public static void main(String[] args) {
		KeyWordComputer kwc = new KeyWordComputer(5);
		String title = " WikiLeaks denied that Snowden had accepted asylum in Venezuela ";
		String content = " There are members of the Russian parliament ,9 No. on social networking site twitter , Snowden, a former CIA employee , Has accepted asylum in Venezuela , But the tweet was deleted a few minutes after it was posted . The Russian authorities declined to comment , Wikileaks, which has been assisting Snowden, denied that he would join Venezuela .　　 Pushkov, chairman of the International Affairs Committee of the Russian parliament , It was first disclosed on personal twitter that Snowden had accepted Venezuela's asylum proposal , Make the outside world think that Snowden's movement has finally made new progress .　　 However, the tweets were immediately deleted within a few minutes , Pushkov clarified that he said this only after he saw the news on Russian state-run television , The TV station has denied it , That pushkov misunderstood the news content .　　 Venezuelan Embassy in Moscow 、 Spokesman of the Russian presidential palace 、 And the foreign ministry declined to comment . WikiLeaks denied that Snowden had officially accepted asylum in Venezuela , Said it would announce the relevant decision at an appropriate time .　　 Snowden believes that he is still at Moscow Sheremetyevo airport , Have been stranded for more than two weeks . He made an appointment with 20 Countries have submitted asylum applications , venezuela 、 Nicaragua and Bolivia , They promised one after another , But Snowden hasn't made a decision yet .　　 And another diplomatic storm , Bolivian President Morales' special plane was refused transit by many European countries last week on the ground of suspecting that Snowden was on the plane , Spain, one of the countries involved, suddenly has a transit wind , Foreign minister margario ] No. 1 expresses its willingness to apologize for any misunderstanding , But he stressed that the authorities did not close the airspace or allow the special plane to land at that time .";
			Collection<Keyword> result = kwc.computeArticleTfidf(title, content);
				System.out.println(result);
		}
	}

Reference resources ：https://github.com/NLPchina/ansj_seg/wiki/%E5%85%B3%E9%94%AE%E8%AF%8D%E6%8A%BD%E5%8F%96https://github.com/NLPchina/ansj_seg/wiki/%E5%85%B3%E9%94%AE%E8%AF%8D%E6%8A%BD%E5%8F%96

summary

This keyword extraction is an incidental feature , It is mainly participle