All Downloads are FREE. Search and download functionalities are using the official Maven repository.
Please wait. This can take some minutes ...
Many resources are needed to download a project. Please understand that we have to compensate our server costs. Thank you in advance.
Project price only 1 $
You can buy this project and download/modify it how often you want.
edu.stanford.nlp.ie.AcquisitionsPrior Maven / Gradle / Ivy
package edu.stanford.nlp.ie;
import java.util.ArrayList;
import java.util.HashSet;
import java.util.List;
import java.util.Set;
import edu.stanford.nlp.util.CoreMap;
import edu.stanford.nlp.util.Index;
import edu.stanford.nlp.util.StringUtils;
/**
* @author Jenny Finkel
*/
public class AcquisitionsPrior extends EntityCachingAbstractSequencePrior {
double penalty = 4.0;
double penalty1 = 3.0;
double penalty2 = 4.0;
public AcquisitionsPrior(String backgroundSymbol, Index classIndex, List doc) {
super(backgroundSymbol, classIndex, doc);
}
public double scoreOf(int[] sequence) {
Set purchasers = new HashSet();
Set purchabrs = new HashSet();
Set sellers = new HashSet();
Set sellerabrs = new HashSet();
Set acquireds = new HashSet();
Set acqabrs = new HashSet();
List purchasersL = new ArrayList();
List purchabrsL = new ArrayList();
List sellersL = new ArrayList();
List sellerabrsL = new ArrayList();
List acquiredsL = new ArrayList();
List acqabrsL = new ArrayList();
double p = 0.0;
for (int i = 0; i < entities.length; i++) {
Entity entity = entities[i];
if ((i == 0 || entities[i-1] != entity) && entity != null) {
String type = classIndex.get(entity.type);
String phrase = StringUtils.join(entity.words, " ").toLowerCase();
if (type.equals("purchaser")) {
purchasers.add(phrase);
purchasersL.add(entity);
} else if (type.equals("purchabr")) {
purchabrs.add(phrase);
purchabrsL.add(entity);
} else if (type.equals("seller")) {
sellers.add(phrase);
sellersL.add(entity);
} else if (type.equals("sellerabr")) {
sellerabrs.add(phrase);
sellerabrsL.add(entity);
} else if (type.equals("acquired")) {
acquireds.add(phrase);
acquiredsL.add(entity);
} else if (type.equals("acqabr")) {
acqabrs.add(phrase);
acqabrsL.add(entity);
} else {
System.err.println("unknown entity type: "+type);
System.exit(0);
}
}
}
for (Entity purchaser : purchasersL) {
if (purchasers.size() > 1) {
p -= purchaser.words.size() * penalty;
}
String s = StringUtils.join(purchaser.words, "").toLowerCase();
boolean match = false;
for (Entity purchabr : purchabrsL) {
String s1 = StringUtils.join(purchabr.words, "").toLowerCase();
//int dist = StringUtils.longestCommonSubstring(s, s1);
//if (dist > s1.length() - 2) {
if (s.indexOf(s1) >= 0) {
match = true;
break;
}
}
if (!match && purchabrs.size() > 0) {
p -= purchaser.words.size() * penalty;
}
}
for (Entity seller : sellersL) {
if (sellers.size() > 1) {
p -= seller.words.size() * penalty;
}
String s = StringUtils.join(seller.words, "").toLowerCase();
boolean match = false;
for (Entity sellerabr : sellerabrsL) {
String s1 = StringUtils.join(sellerabr.words, "").toLowerCase();
//int dist = StringUtils.longestCommonSubstring(s, s1);
//if (dist > s1.length() - 2) {
if (s.indexOf(s1) >= 0) {
match = true;
break;
}
}
if (!match && sellerabrs.size() > 0) {
p -= seller.words.size() * penalty;
}
}
for (Entity acquired : acquiredsL) {
if (acquireds.size() > 1) {
p -= acquired.words.size() * penalty;
}
String s = StringUtils.join(acquired.words, "").toLowerCase();
boolean match = false;
for (Entity acqabr : acqabrsL) {
String s1 = StringUtils.join(acqabr.words, "").toLowerCase();
//int dist = StringUtils.longestCommonSubstring(s, s1);
//if (dist > s1.length() - 2) {
if (s.indexOf(s1) >= 0) {
match = true;
break;
}
}
if (!match && acqabrs.size() > 0) {
p -= acquired.words.size() * penalty;
}
}
for (Entity purchabr : purchabrsL) {
//p -= purchabr.words.size() * penalty;
String s = StringUtils.join(purchabr.words, "").toLowerCase();
boolean match = false;
for (Entity purchaser : purchasersL) {
String s1 = StringUtils.join(purchaser.words, "").toLowerCase();
//int dist = StringUtils.longestCommonSubstring(s, s1);
//if (dist > s1.length() - 2) {
if (s1.indexOf(s) >= 0) {
match = true;
break;
}
}
if (!match) {
p -= purchabr.words.size() * penalty2;
}
match = false;
for (Entity acquired : acquiredsL) {
String s1 = StringUtils.join(acquired.words, "").toLowerCase();
//int dist = StringUtils.longestCommonSubstring(s, s1);
//if (dist > s.length() - 2) {
if (s1.indexOf(s) >= 0) {
match = true;
break;
}
}
for (Entity seller : sellersL) {
String s1 = StringUtils.join(seller.words, "").toLowerCase();
//int dist = StringUtils.longestCommonSubstring(s, s1);
//if (dist > s.length() - 2) {
if (s1.indexOf(s) >= 0) {
match = true;
break;
}
}
if (match) {
p -= purchabr.words.size() * penalty1;
}
}
for (Entity sellerabr : sellerabrsL) {
//p -= sellerabr.words.size() * penalty;
String s = StringUtils.join(sellerabr.words, "").toLowerCase();
boolean match = false;
for (Entity seller : sellersL) {
String s1 = StringUtils.join(seller.words, "").toLowerCase();
//int dist = StringUtils.longestCommonSubstring(s, s1);
//if (dist > s1.length() - 2) {
if (s1.indexOf(s) >= 0) {
match = true;
break;
}
}
if (!match) {
p -= sellerabr.words.size() * penalty2;
}
match = false;
for (Entity acquired : acquiredsL) {
String s1 = StringUtils.join(acquired.words, "").toLowerCase();
//int dist = StringUtils.longestCommonSubstring(s, s1);
//if (dist > s.length() - 2) {
if (s1.indexOf(s) >= 0) {
match = true;
break;
}
}
for (Entity purchaser : purchasersL) {
String s1 = StringUtils.join(purchaser.words, "").toLowerCase();
//int dist = StringUtils.longestCommonSubstring(s, s1);
//if (dist > s.length() - 2) {
if (s1.indexOf(s) >= 0) {
match = true;
break;
}
}
if (match) {
p -= sellerabr.words.size() * penalty1;
}
}
for (Entity acqabr : acqabrsL) {
//p -= acqabr.words.size() * penalty;
String s = StringUtils.join(acqabr.words, "").toLowerCase();
boolean match = false;
for (Entity acquired : acquiredsL) {
String s1 = StringUtils.join(acquired.words, "").toLowerCase();
//int dist = StringUtils.longestCommonSubstring(s, s1);
//if (dist > s1.length() - 2) {
if (s1.indexOf(s) >= 0) {
match = true;
break;
}
}
if (!match) {
p -= acqabr.words.size() * penalty2;
}
match = false;
for (Entity seller : sellersL) {
String s1 = StringUtils.join(seller.words, "").toLowerCase();
//int dist = StringUtils.longestCommonSubstring(s, s1);
//if (dist > s.length() - 2) {
if (s1.indexOf(s) >= 0) {
//System.err.println(acqabr.toString(classIndex)+"\n"+seller.toString(classIndex)+"\n");
match = true;
break;
}
}
for (Entity purchaser : purchasersL) {
String s1 = StringUtils.join(purchaser.words, "").toLowerCase();
//int dist = StringUtils.longestCommonSubstring(s, s1);
//if (dist > s.length() - 2) {
if (s1.indexOf(s) >= 0) {
match = true;
break;
}
}
if (match) {
p -= acqabr.words.size() * penalty1;
}
}
return p;
}
}