smile.nlp.collocation

Class Bigram

java.lang.Object

smile.nlp.Bigram

smile.nlp.collocation.Bigram

All Implemented Interfaces:

java.lang.Comparable<Bigram>

public class Bigram
extends Bigram
implements java.lang.Comparable<Bigram>

Collocations are expressions of multiple words which commonly co-occur. A bigram collocation is a pair of words w1 w2 that appear together with statistically significance.

Field Summary

Fields

Modifier and Type	Field and Description
int	count The frequency of bigram in the corpus.
double	score The chi-square statistical score of the collocation.

Fields inherited from class smile.nlp.Bigram

w1, w2

Constructor Summary

Constructors

Constructor and Description
Bigram(java.lang.String w1, java.lang.String w2, int count, double score) Constructor.

Method Summary

Modifier and Type	Method and Description
int	compareTo(Bigram o)
static Bigram[]	of(Corpus corpus, double p, int minFrequency) Finds bigram collocations in the given corpus whose p-value is less than the given threshold.
static Bigram[]	of(Corpus corpus, int k, int minFrequency) Finds top k bigram collocations in the given corpus.
java.lang.String	toString()

Methods inherited from class smile.nlp.Bigram

equals, hashCode

Methods inherited from class java.lang.Object

clone, finalize, getClass, notify, notifyAll, wait, wait, wait

Field Detail

count

public final int count

The frequency of bigram in the corpus.

score

public final double score

The chi-square statistical score of the collocation.

Constructor Detail

Bigram

public Bigram(java.lang.String w1,
              java.lang.String w2,
              int count,
              double score)

Constructor.

Parameters:

w1 - the first word of bigram.

w2 - the second word of bigram.

count - the frequency of bigram in the corpus.

score - the chi-square statistical score of collocation in a corpus.

Method Detail

toString

public java.lang.String toString()

Overrides:

toString in class Bigram

compareTo

public int compareTo(Bigram o)

Specified by:

compareTo in interface java.lang.Comparable<Bigram>

of

public static Bigram[] of(Corpus corpus,
                          int k,
                          int minFrequency)

Finds top k bigram collocations in the given corpus.

Parameters:

minFrequency - The minimum frequency of bigram in the corpus.

Returns:

the array of significant bigram collocations in descending order of likelihood ratio.

of

public static Bigram[] of(Corpus corpus,
                          double p,
                          int minFrequency)

Finds bigram collocations in the given corpus whose p-value is less than the given threshold.

Parameters:

p - the p-value threshold

minFrequency - The minimum frequency of bigram in the corpus.

Returns:

the array of significant bigram collocations in descending order of likelihood ratio.