edu.illinois.cs.cogcomp.lbj.coref.ir.examples
Class WordExample

java.lang.Object
  edu.illinois.cs.cogcomp.lbj.coref.ir.examples.Example
      edu.illinois.cs.cogcomp.lbj.coref.ir.examples.WordExample

Direct Known Subclasses:

BIOExample, ExtendHeadExample

public class WordExample
extends Example

Represents a word in a document, for use in classification or classifier training, along with utility methods for getting information and features pertaining to the word.

Field Summary

protected Doc	m_doc The document containing the example word.
protected int	m_numWords The number of words in the document.
protected int	m_wordN The (zero-based) position of the example word in the document.

Constructor Summary

WordExample()
Constructs an example that does not refer to any word.

WordExample(Doc doc, int wordNum)
Constructs a word example from a given document and word number.

Method Summary

java.lang.String	getCasedWord() Gets the word at the word number plus the offset, without altering original case; Note this does NOT mean that any extra casing is applied here (although it may be applied at the Doc parsing stage).
java.lang.String	getCasedWord(int offset) Gets the word without altering original case; Note this does NOT mean that any extra casing is applied here (although it may be applied at the Doc parsing stage.)
Doc	getDoc()
java.lang.String	getPOS() Gets the part-of-speech (POS) tag of the word represented by this example.
java.lang.String	getPOS(int offset) Gets the part-of-speech (POS) tag of the word at the word number plus the offset.
java.lang.String	getWord() Gets a lowercase version of the word.
java.lang.String	getWord(int offset) Gets a lowercase version of the word at the word number plus the offset, or the empty string if no such word exists.
java.lang.String	getWordCase() Gets a description of the case of the word.
java.lang.String	getWordCase(int offset) Gets a description of the case of the specified word.
java.lang.String	getWordCase(java.lang.String word) Gets a description of the case of the specified word.
int	getWordNum()

Methods inherited from class edu.illinois.cs.cogcomp.lbj.coref.ir.examples.Example

getLabel

Methods inherited from class java.lang.Object

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

Field Detail

m_doc

protected Doc m_doc

The document containing the example word.

m_wordN

protected int m_wordN

The (zero-based) position of the example word in the document.

m_numWords

protected int m_numWords

The number of words in the document.

Constructor Detail

WordExample

public WordExample()

Constructs an example that does not refer to any word. Not recommended for use unless word will be set.

WordExample

public WordExample(Doc doc,
                   int wordNum)

Constructs a word example from a given document and word number.

Parameters:

doc - The document containing the example word.

wordNum - The (zero-based) position of the word in the document.

Method Detail

getDoc

public Doc getDoc()

getWordNum

public int getWordNum()

getWord

public java.lang.String getWord()

Gets a lowercase version of the word.

Returns:

the word, in lowercase.

getWord

public java.lang.String getWord(int offset)

Gets a lowercase version of the word at the word number plus the offset, or the empty string if no such word exists.

Parameters:

offset - The position relative to the word number of the desired word.

Returns:

the word, in lowercase, or the empty string if no such word.

getCasedWord

public java.lang.String getCasedWord()

Gets the word at the word number plus the offset, without altering original case; Note this does NOT mean that any extra casing is applied here (although it may be applied at the Doc parsing stage).

Returns:

The word (without altering its case).

getCasedWord

public java.lang.String getCasedWord(int offset)

Gets the word without altering original case; Note this does NOT mean that any extra casing is applied here (although it may be applied at the Doc parsing stage.)

Parameters:

offset - The position relative to the word number of the desired word.

Returns:

The word (without altering its case), or the empty string if no such word exists.

getPOS

public java.lang.String getPOS()

Gets the part-of-speech (POS) tag of the word represented by this example.

Returns:

The part-of-speech tag of the word.

See Also:

Doc.getPOS(int)

getPOS

public java.lang.String getPOS(int offset)

Gets the part-of-speech (POS) tag of the word at the word number plus the offset.

Parameters:

offset - The position relative to the word number of the desired word's POS tag.

Returns:

The part-of-speech tag of the specified word.

See Also:

Doc.getPOS(int)

getWordCase

public java.lang.String getWordCase()

Gets a description of the case of the word.

Returns:

"allLower", "firstCap", "allCaps", "multiCase", "digit", "punc", "other" In case of a single-character uppercase word, returns "firstCap" Words beginning with a digit are "digit". This implies that words containing internal digits can still be "allLower" or "allUpper". Words beginning with punctuation are "punc", and word-internal punc is not considered "punc". Zero length words are "other". Words beginning with whitespace are "other". multiCase means mixed case, but "mixed*" as a feature is disallowed.

getWordCase

public java.lang.String getWordCase(int offset)

Gets a description of the case of the specified word.

Parameters:

offset - The position relative to the word number of the specified word.

Returns:

"allLower", "firstCap", "allCaps", "multiCase", "digit", "punc", "other" In case of a single-character uppercase word, returns "firstCap" Words beginning with a digit are "digit". This implies that words containing internal digits can still be "allLower" or "allUpper". Words beginning with punctuation are "punc", and word-internal punc is not considered "punc". Zero length words are "other". Words beginning with whitespace are "other". multiCase means mixed case, but "mixed*" as a feature is disallowed.

getWordCase

public java.lang.String getWordCase(java.lang.String word)

Gets a description of the case of the specified word.

Parameters:

word - The specified word.

Returns:

"allLower", "firstCap", "allCaps", "multiCase", "digit", "punc", "other" In case of a single-character uppercase word, returns "firstCap" Words beginning with a digit are "digit". This implies that words containing internal digits can still be "allLower" or "allUpper". Words beginning with punctuation are "punc", and word-internal punc is not considered "punc". Zero length words are "other". Words beginning with whitespace are "other". multiCase means mixed case, but "mixed*" as a feature is disallowed.