edu.umass.cs.mallet.projects.seg_plus_coref.coreference

Class CitationUtils

java.lang.Object

edu.umass.cs.mallet.projects.seg_plus_coref.coreference.CitationUtils

public class CitationUtils
extends Object

Field Summary

Fields

Modifier and Type	Field and Description
static String	AUTHOR
static String	PAPER
static String[]	SEPERATOR
static String	VENUE

Constructor Summary

Constructors

Constructor and Description
CitationUtils()

Method Summary

Modifier and Type	Method and Description
static void	addAuthorsFromLine(String str, ArrayList nodes, IEInterface ieInterface, boolean useCRFLocal, int numNBest, int nthViterbi)
static void	addPaperFromLine(String str, ArrayList nodes, IEInterface ieInterface, boolean useCRFLocal, int numNBest, int nthViterbi)
static void	addVenuesFromLine(String str, ArrayList nodes, IEInterface ieInterface, boolean useCRFLocal, int numNBest, int nthViterbi)
static com.wcohen.secondstring.StringDistance	computeDistanceMetric(ArrayList nodes)
static ArrayList	computeNodes(ArrayList trainFileArray, IEInterface ieInterface, boolean useCRFLocal, int numNBest, int nthViterbi)
static ArrayList	computeNodes(ArrayList trainFileArray, IEInterface ieInterface, boolean useCRFLocal, int numNBest, int nthViterbi, String nodeType) Build nodes with labels corresponding to cluster ids.
protected static ArrayList	computeNodesWPubs(ArrayList trainFileArray, ArrayList publications, IEInterface ieInterface, boolean useCRFLocal, int numNBest, int nthViterbi)
static ArrayList	computeNodesWPubs(ArrayList trainFileArray, ArrayList publications, IEInterface ieInterface, int numNBest, int nthViterbi)
static void	evaluateClustering(Collection key, Collection pred, String description)
static Collection	makeCollections(ArrayList nodes)
static void	makeDistMetric(List list, com.wcohen.secondstring.StringDistance tfidf, com.wcohen.secondstring.StringDistance triGramDistanceMetric)
static InstanceList	makePairs(Pipe instancePipe, ArrayList nodes)
static InstanceList	makePairs(Pipe instancePipe, ArrayList nodes, double negativeProb)
static InstanceList	makePairs(Pipe instancePipe, ArrayList nodes, List pairs)

Methods inherited from class java.lang.Object

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

Field Detail

PAPER

public static final String PAPER

See Also:

Constant Field Values

VENUE

public static final String VENUE

See Also:

Constant Field Values

AUTHOR

public static final String AUTHOR

See Also:

Constant Field Values

SEPERATOR

public static String[] SEPERATOR

Constructor Detail

CitationUtils

public CitationUtils()

Method Detail

computeNodes

public static ArrayList computeNodes(ArrayList trainFileArray,
                                     IEInterface ieInterface,
                                     boolean useCRFLocal,
                                     int numNBest,
                                     int nthViterbi)

addPaperFromLine

public static void addPaperFromLine(String str,
                                    ArrayList nodes,
                                    IEInterface ieInterface,
                                    boolean useCRFLocal,
                                    int numNBest,
                                    int nthViterbi)

addVenuesFromLine

public static void addVenuesFromLine(String str,
                                     ArrayList nodes,
                                     IEInterface ieInterface,
                                     boolean useCRFLocal,
                                     int numNBest,
                                     int nthViterbi)

addAuthorsFromLine

public static void addAuthorsFromLine(String str,
                                      ArrayList nodes,
                                      IEInterface ieInterface,
                                      boolean useCRFLocal,
                                      int numNBest,
                                      int nthViterbi)

computeNodes

public static ArrayList computeNodes(ArrayList trainFileArray,
                                     IEInterface ieInterface,
                                     boolean useCRFLocal,
                                     int numNBest,
                                     int nthViterbi,
                                     String nodeType)

Build nodes with labels corresponding to cluster ids. Passing different strings for nodeType will make node correspond to the paper, author, or venue, etc. Default is paper.

Parameters:

trainFileArray - list of citation files

ieInterface - the interface to the extraction

useCRFLocal - use a CRF for segmentation

numNBest - number of viterbi paths to use

nodeType - type of node to create (paper, author, venue...)

computeNodesWPubs

public static ArrayList computeNodesWPubs(ArrayList trainFileArray,
                                          ArrayList publications,
                                          IEInterface ieInterface,
                                          int numNBest,
                                          int nthViterbi)

computeNodesWPubs

protected static ArrayList computeNodesWPubs(ArrayList trainFileArray,
                                             ArrayList publications,
                                             IEInterface ieInterface,
                                             boolean useCRFLocal,
                                             int numNBest,
                                             int nthViterbi)

computeDistanceMetric

public static com.wcohen.secondstring.StringDistance computeDistanceMetric(ArrayList nodes)

makeDistMetric

public static void makeDistMetric(List list,
                                  com.wcohen.secondstring.StringDistance tfidf,
                                  com.wcohen.secondstring.StringDistance triGramDistanceMetric)

makePairs

public static InstanceList makePairs(Pipe instancePipe,
                                     ArrayList nodes)

makePairs

public static InstanceList makePairs(Pipe instancePipe,
                                     ArrayList nodes,
                                     double negativeProb)

Parameters:

instancePipe - to pipe instances through

nodes - Citation nodes

negativeProb - the probability of including each negative instance (to reduce class disparity)

makePairs

public static InstanceList makePairs(Pipe instancePipe,
                                     ArrayList nodes,
                                     List pairs)

makeCollections

public static Collection makeCollections(ArrayList nodes)

evaluateClustering

public static void evaluateClustering(Collection key,
                                      Collection pred,
                                      String description)