de.aitools.ie.keyphraseextraction
Class KeyphraseExtractor

java.lang.Object
  de.aitools.ie.keyphraseextraction.KeyphraseExtractor

Direct Known Subclasses:

FrequentPhraseExtractor, HeadNounPhraseExtractor, RepeatedStringExtractor, StanfordNounPhraseExtractor, TextRankExtractor

public abstract class KeyphraseExtractor
extends java.lang.Object

An abstract class to extract key phrases from given text.

Version:

$Id: KeyphraseExtractor.java,v 1.8 2011/12/15 15:09:07 hoppe Exp $

Author:

martin.trenkmann@uni-weimar.de

See Also:

HeadNounPhraseExtractor, RepeatedStringExtractor

Method Summary

java.util.SortedSet<Phrase>	extract(java.lang.String text) Extracts all key phrases from the given text with score normalization to the range of [0, 1].
java.util.SortedSet<Phrase>	extract(java.lang.String text, int k) Extracts at most the k highest rated key phrases from the given text with score normalization to the range of [0, 1].
java.util.SortedSet<Phrase>	extract(java.lang.String text, int k, boolean normalize) Extracts at most the k highest rated key phrases from the given text with or without score normalization to the range of [0, 1].
java.util.SortedSet<Phrase>	extract(java.lang.String text, int k, int phraseLength, boolean normalize) Extracts at most the k highest rated key phrases from the given text with or without score normalization to the range of [0, 1].
java.util.Locale	getLocale() Returns the current locale.
static java.util.SortedSet<Phrase>	getTopPhrases(java.util.Set<Phrase> phrases, int k) Extracts the k highest rated phrases.
void	setLocale(java.util.Locale locale) Registers a new locale to this extractor.

Methods inherited from class java.lang.Object

equals, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Method Detail

extract

public java.util.SortedSet<Phrase> extract(java.lang.String text)

Extracts all key phrases from the given text with score normalization to the range of [0, 1].

Parameters:

text - the text to extract key phrases from.

Returns:

a ranked set of the highest scored key phrases.

See Also:

extract(String, int, int, boolean)

extract

public java.util.SortedSet<Phrase> extract(java.lang.String text,
                                           int k)

Extracts at most the k highest rated key phrases from the given text with score normalization to the range of [0, 1].

Parameters:

text - the text to extract key phrases from.

k - the maximal number of key phrases to extract.

Returns:

a ranked set of the highest scored key phrases.

See Also:

extract(String, int, int, boolean)

extract

public java.util.SortedSet<Phrase> extract(java.lang.String text,
                                           int k,
                                           boolean normalize)

Extracts at most the k highest rated key phrases from the given text with or without score normalization to the range of [0, 1].

Parameters:

text - the text to extract key phrases from.

k - the maximal number of key phrases to extract.

normalize - enables/disables score normalization.

Returns:

a ranked set of the highest scored key phrases.

See Also:

extract(String, int, int, boolean)

extract

public java.util.SortedSet<Phrase> extract(java.lang.String text,
                                           int k,
                                           int phraseLength,
                                           boolean normalize)

Extracts at most the k highest rated key phrases from the given text with or without score normalization to the range of [0, 1]. All phrases can be limited in their word-length by phraseLength.

Parameters:

text - the text to extract key phrases from.

k - the maximal number of key phrases to extract.

phraseLength - the maximal length of key phrases in words.

normalize - enables/disables score normalization to the range [0,1].

Returns:

a ranked set of the highest scored key phrases.

getLocale

public java.util.Locale getLocale()

Returns the current locale. The locale parameter determines how to tag words with its part-of-speech and how to recognize stopwords.

Returns:

the current locale.

setLocale

public void setLocale(java.util.Locale locale)

Registers a new locale to this extractor. This method may be overwritten in inherited classes.

Parameters:

locale - the locale to register.

getTopPhrases

public static java.util.SortedSet<Phrase> getTopPhrases(java.util.Set<Phrase> phrases,
                                                        int k)

Extracts the k highest rated phrases. If the set of given phrases contains less than k elements, the entire set is returned in a sorted form.

Parameters:

phrases - a set of phrases.

k - the maximal number of phrases to extract.

Returns:

a sorted set of at most k phrases.