org.apache.ctakes.ytex.kernel.pagerank

Class PageRankServiceImpl

java.lang.Object

org.apache.ctakes.ytex.kernel.pagerank.PageRankServiceImpl

All Implemented Interfaces:

PageRankService

public class PageRankServiceImpl
extends Object
implements PageRankService

Constructor Summary

Constructors

Constructor and Description
PageRankServiceImpl()

Method Summary

Modifier and Type	Method and Description
static void	main(String[] args)
double[]	pagerankIter(double[] currentScores, Map<Integer,Double> dampingVector, ConceptGraph cg, double dampingFactor, double N)
double[]	pagerankIter(double[] currentScores, Map<Integer,Double> dampingVector, ConceptGraph cg, double dampingFactor, double N, Set<Integer> activeNodes)
Map<Integer,Double>	pagerankIter(Map<Integer,Double> currentScores, Map<Integer,Double> dampingVector, ConceptGraph cg, double dampingFactor, double N) perform one iteration of pagerank
double[]	rank(Map<String,Double> dampingVector, ConceptGraph cg) call rank() with default values for iter (30), threshold(1e-4), dampingFactor(0.85)
double[]	rank(Map<String,Double> dampingVector, ConceptGraph cg, int iter, double threshold, double dampingFactor) PageRank for conceptGraph.
double[]	rank2(Map<Integer,Double> dampingVector, ConceptGraph cg, int iter, double threshold, double dampingFactor)
double	sim(String concept1, String concept2, ConceptGraph cg, int iter, double threshold, double dampingFactor) compute similarity using personalized page rank, as documented in Exploring Knowledge Bases for Similarity

Methods inherited from class java.lang.Object

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

Constructor Detail

PageRankServiceImpl

public PageRankServiceImpl()

Method Detail

pagerankIter

public double[] pagerankIter(double[] currentScores,
                             Map<Integer,Double> dampingVector,
                             ConceptGraph cg,
                             double dampingFactor,
                             double N,
                             Set<Integer> activeNodes)

pagerankIter

public double[] pagerankIter(double[] currentScores,
                             Map<Integer,Double> dampingVector,
                             ConceptGraph cg,
                             double dampingFactor,
                             double N)

rank2

public double[] rank2(Map<Integer,Double> dampingVector,
                      ConceptGraph cg,
                      int iter,
                      double threshold,
                      double dampingFactor)

Specified by:

rank2 in interface PageRankService

pagerankIter

public Map<Integer,Double> pagerankIter(Map<Integer,Double> currentScores,
                                        Map<Integer,Double> dampingVector,
                                        ConceptGraph cg,
                                        double dampingFactor,
                                        double N)

perform one iteration of pagerank

Parameters:

currentScores -

cg -

Returns:

sim

public double sim(String concept1,
                  String concept2,
                  ConceptGraph cg,
                  int iter,
                  double threshold,
                  double dampingFactor)

compute similarity using personalized page rank, as documented in Exploring Knowledge Bases for Similarity

Specified by:

sim in interface PageRankService

Parameters:

concept1 -

concept2 -

cg -

iter -

threshold -

dampingFactor -

Returns:

main

public static void main(String[] args)

rank

public double[] rank(Map<String,Double> dampingVector,
                     ConceptGraph cg,
                     int iter,
                     double threshold,
                     double dampingFactor)

Description copied from interface: PageRankService

PageRank for conceptGraph. Page = concept. in-links = parents. out-links = children.

Specified by:

rank in interface PageRankService

Parameters:

dampingVector - topic vector/personalized pagerank vector. If null will use normal pagerank with a damping vector where every value is 1/N

cg - concept graph

iter - max number of iterations

threshold - convergence threshold

Returns:

pageRank 'vector'. key = concept (page), value = rank

rank

public double[] rank(Map<String,Double> dampingVector,
                     ConceptGraph cg)

Description copied from interface: PageRankService

call rank() with default values for iter (30), threshold(1e-4), dampingFactor(0.85)

Specified by:

rank in interface PageRankService

Returns: