update in the discovery of clustering, conditions and distance functions (annotated with custom annotations)

This commit is contained in:
Michele De Bonis 2018-10-24 12:09:41 +02:00
parent 4d379c2227
commit 1cbbc3f15a
62 changed files with 441 additions and 204 deletions

View File

@ -55,6 +55,12 @@
<artifactId>junit</artifactId>
<scope>test</scope>
</dependency>
<dependency>
<groupId>org.reflections</groupId>
<artifactId>reflections</artifactId>
<version>0.9.10</version>
</dependency>
</dependencies>
</project>

View File

@ -18,7 +18,13 @@ public abstract class AbstractClusteringFunction extends AbstractPaceFunctions i
public AbstractClusteringFunction(final Map<String, Integer> params) {
this.params = params;
}
public AbstractClusteringFunction(){}
public void setParams(Map<String, Integer> params){
this.params = params;
}
protected abstract Collection<String> doApply(String s);
@Override

View File

@ -7,12 +7,17 @@ import java.util.StringTokenizer;
import com.google.common.collect.Sets;
@ClusteringClass("acronyms")
public class Acronyms extends AbstractClusteringFunction {
public Acronyms(Map<String, Integer> params) {
super(params);
}
public Acronyms(){
super();
}
@Override
protected Collection<String> doApply(String s) {
return extractAcronyms(s, param("max"), param("minLen"), param("maxLen"));

View File

@ -22,9 +22,6 @@ public class BlacklistAwareClusteringCombiner extends ClusteringCombiner {
private static final Log log = LogFactory.getLog(BlacklistAwareClusteringCombiner.class);
public static Collection<String> filterAndCombine(final MapDocument a, final Config conf) {
final Document filtered = new BlacklistAwareClusteringCombiner().filter(a, conf.blacklists());

View File

@ -1,5 +0,0 @@
package eu.dnetlib.pace.clustering;
public enum Clustering {
acronyms, ngrams, ngrampairs, sortedngrampairs, suffixprefix, spacetrimmingfieldvalue, immutablefieldvalue, personhash, personclustering, lowercase, urlclustering
}

View File

@ -0,0 +1,13 @@
package eu.dnetlib.pace.clustering;
import java.lang.annotation.ElementType;
import java.lang.annotation.Retention;
import java.lang.annotation.RetentionPolicy;
import java.lang.annotation.Target;
@Retention(RetentionPolicy.RUNTIME)
@Target(ElementType.TYPE)
public @interface ClusteringClass {
public String value();
}

View File

@ -12,4 +12,5 @@ public interface ClusteringFunction {
public Map<String, Integer> getParams();
public void setParams(Map<String, Integer> params);
}

View File

@ -0,0 +1,24 @@
package eu.dnetlib.pace.clustering;
import java.io.Serializable;
import java.lang.reflect.InvocationTargetException;
import java.util.Map;
import java.util.stream.Collectors;
import org.reflections.Reflections;
public class ClusteringResolver implements Serializable {
private final Map<String, Class<ClusteringFunction>> functionMap;
public ClusteringResolver() {
this.functionMap = new Reflections("eu.dnetlib").getTypesAnnotatedWith(ClusteringClass.class).stream()
.filter(ClusteringFunction.class::isAssignableFrom)
.collect(Collectors.toMap(cl -> cl.getAnnotation(ClusteringClass.class).value(), cl -> (Class<ClusteringFunction>)cl));
}
public ClusteringFunction resolve(String clusteringFunction) throws IllegalAccessException, InstantiationException, NoSuchMethodException, InvocationTargetException {
return functionMap.get(clusteringFunction).newInstance();
}
}

View File

@ -6,12 +6,17 @@ import java.util.Map;
import com.google.common.collect.Lists;
@ClusteringClass("immutablefieldvalue")
public class ImmutableFieldValue extends AbstractClusteringFunction {
public ImmutableFieldValue(final Map<String, Integer> params) {
super(params);
}
public ImmutableFieldValue() {
super();
}
@Override
protected Collection<String> doApply(final String s) {
final List<String> res = Lists.newArrayList();

View File

@ -9,12 +9,17 @@ import com.google.common.collect.Sets;
import eu.dnetlib.pace.model.Field;
import org.apache.commons.lang.StringUtils;
@ClusteringClass("lowercase")
public class LowercaseClustering extends AbstractClusteringFunction {
public LowercaseClustering(final Map<String, Integer> params) {
super(params);
}
public LowercaseClustering(){
super();
}
@Override
public Collection<String> apply(List<Field> fields) {
Collection<String> c = Sets.newLinkedHashSet();

View File

@ -6,8 +6,13 @@ import java.util.Map;
import com.google.common.collect.Lists;
@ClusteringClass("ngrampairs")
public class NgramPairs extends Ngrams {
public NgramPairs() {
super();
}
public NgramPairs(Map<String, Integer> params) {
super(params);
}

View File

@ -5,12 +5,17 @@ import java.util.LinkedHashSet;
import java.util.Map;
import java.util.StringTokenizer;
@ClusteringClass("ngrams")
public class Ngrams extends AbstractClusteringFunction {
public Ngrams(Map<String, Integer> params) {
super(params);
}
public Ngrams() {
super();
}
@Override
protected Collection<String> doApply(String s) {
return getNgrams(s, param("ngramLen"), param("max"), param("maxPerToken"), param("minNgramLen"));

View File

@ -1,5 +1,6 @@
package eu.dnetlib.pace.clustering;
import java.io.Serializable;
import java.util.Collection;
import java.util.List;
import java.util.Map;
@ -18,6 +19,7 @@ import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.model.gt.Author;
import eu.dnetlib.pace.model.gt.GTAuthor;
@ClusteringClass("personclustering")
public class PersonClustering extends AbstractPaceFunctions implements ClusteringFunction {
private Map<String, Integer> params;
@ -28,6 +30,10 @@ public class PersonClustering extends AbstractPaceFunctions implements Clusterin
this.params = params;
}
public void setParams(Map<String, Integer> params){
this.params = params;
}
@Override
public Collection<String> apply(final List<Field> fields) {
final Set<String> hashes = Sets.newHashSet();

View File

@ -8,6 +8,7 @@ import com.google.common.collect.Lists;
import eu.dnetlib.pace.model.Person;
@ClusteringClass("personhash")
public class PersonHash extends AbstractClusteringFunction {
private boolean DEFAULT_AGGRESSIVE = false;
@ -16,6 +17,10 @@ public class PersonHash extends AbstractClusteringFunction {
super(params);
}
public PersonHash(){
super();
}
@Override
protected Collection<String> doApply(final String s) {
final List<String> res = Lists.newArrayList();

View File

@ -9,6 +9,10 @@ public class RandomClusteringFunction extends AbstractClusteringFunction {
super(params);
}
public RandomClusteringFunction(){
super();
}
@Override
protected Collection<String> doApply(String s) {
// TODO Auto-generated method stub

View File

@ -9,12 +9,17 @@ import com.google.common.base.Joiner;
import com.google.common.base.Splitter;
import com.google.common.collect.Lists;
@ClusteringClass("sortedngrampairs")
public class SortedNgramPairs extends NgramPairs {
public SortedNgramPairs(Map<String, Integer> params) {
super(params);
}
public SortedNgramPairs(){
super();
}
@Override
protected Collection<String> doApply(String s) {

View File

@ -9,12 +9,17 @@ import org.apache.commons.lang.StringUtils;
import com.google.common.collect.Lists;
@ClusteringClass("spacetrimmingfieldvalue")
public class SpaceTrimmingFieldValue extends AbstractClusteringFunction {
public SpaceTrimmingFieldValue(final Map<String, Integer> params) {
super(params);
}
public SpaceTrimmingFieldValue(){
super();
}
@Override
protected Collection<String> doApply(final String s) {
final List<String> res = Lists.newArrayList();

View File

@ -6,12 +6,17 @@ import java.util.Set;
import com.google.common.collect.Sets;
@ClusteringClass("suffixprefix")
public class SuffixPrefix extends AbstractClusteringFunction {
public SuffixPrefix(Map<String, Integer> params) {
super(params);
}
public SuffixPrefix(){
super();
}
@Override
protected Collection<String> doApply(String s) {
return suffixPrefix(s, param("len"), param("max"));

View File

@ -11,6 +11,7 @@ import java.util.List;
import java.util.Map;
import java.util.stream.Collectors;
@ClusteringClass("urlclustering")
public class UrlClustering extends AbstractPaceFunctions implements ClusteringFunction {
protected Map<String, Integer> params;
@ -19,6 +20,14 @@ public class UrlClustering extends AbstractPaceFunctions implements ClusteringFu
this.params = params;
}
public UrlClustering() {
super();
}
public void setParams(Map<String, Integer> params){
this.params = params;
}
@Override
public Collection<String> apply(List<Field> fields) {
return fields.stream()

View File

@ -2,7 +2,6 @@ package eu.dnetlib.pace.condition;
import java.util.List;
import eu.dnetlib.pace.common.AbstractPaceFunctions;
import eu.dnetlib.pace.config.Cond;
import eu.dnetlib.pace.distance.eval.ConditionEval;
import eu.dnetlib.pace.distance.eval.ConditionEvalMap;
import eu.dnetlib.pace.model.Document;
@ -17,15 +16,25 @@ import eu.dnetlib.pace.model.FieldDef;
*/
public abstract class AbstractCondition extends AbstractPaceFunctions implements ConditionAlgo {
protected Cond cond;
protected String cond;
protected List<FieldDef> fields;
public AbstractCondition(final Cond cond, final List<FieldDef> fields) {
public AbstractCondition(final String cond, final List<FieldDef> fields) {
this.cond = cond;
this.fields = fields;
}
public AbstractCondition(){}
public void setCond(String cond){
this.cond = cond;
}
public void setFields(List<FieldDef> fields){
this.fields = fields;
}
protected abstract ConditionEval verify(FieldDef fd, Field a, Field b);
@Override

View File

@ -1,7 +1,6 @@
package eu.dnetlib.pace.condition;
import java.util.List;
import eu.dnetlib.pace.config.Cond;
import eu.dnetlib.pace.distance.eval.ConditionEval;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.model.FieldDef;
@ -11,12 +10,16 @@ import eu.dnetlib.pace.model.FieldDef;
*
* @author claudio
*/
@ConditionClass("alwaystruecondition")
public class AlwaysTrueCondition extends AbstractCondition {
public AlwaysTrueCondition(final Cond cond, final List<FieldDef> fields) {
public AlwaysTrueCondition(final String cond, final List<FieldDef> fields) {
super(cond, fields);
}
public AlwaysTrueCondition(){
super();
}
@Override
protected ConditionEval verify(final FieldDef fd, final Field a, final Field b) {
return new ConditionEval(cond, a, b, 1);

View File

@ -1,9 +1,9 @@
package eu.dnetlib.pace.condition;
import java.util.Map;
import java.util.List;
import eu.dnetlib.pace.distance.eval.ConditionEvalMap;
import eu.dnetlib.pace.model.Document;
import eu.dnetlib.pace.model.FieldDef;
/**
* Allows to express general conditions to be satisfied or not between two Documents.
@ -24,4 +24,7 @@ public interface ConditionAlgo {
*/
public abstract ConditionEvalMap verify(Document a, Document b);
public void setFields(List<FieldDef> fields);
public void setCond(String name);
}

View File

@ -0,0 +1,13 @@
package eu.dnetlib.pace.condition;
import java.lang.annotation.ElementType;
import java.lang.annotation.Retention;
import java.lang.annotation.RetentionPolicy;
import java.lang.annotation.Target;
@Retention(RetentionPolicy.RUNTIME)
@Target(ElementType.TYPE)
public @interface ConditionClass {
public String value();
}

View File

@ -0,0 +1,22 @@
package eu.dnetlib.pace.condition;
import java.io.Serializable;
import java.util.Map;
import java.util.stream.Collectors;
import org.reflections.Reflections;
public class ConditionResolver implements Serializable {
private final Map<String, Class<ConditionAlgo>> functionMap;
public ConditionResolver() {
this.functionMap = new Reflections("eu.dnetlib").getTypesAnnotatedWith(ConditionClass.class).stream()
.filter(ConditionAlgo.class::isAssignableFrom)
.collect(Collectors.toMap(cl -> cl.getAnnotation(ConditionClass.class).value(), cl -> (Class<ConditionAlgo>)cl));
}
public ConditionAlgo resolve(String name) throws IllegalAccessException, InstantiationException {
return functionMap.get(name).newInstance();
}
}

View File

@ -2,7 +2,6 @@ package eu.dnetlib.pace.condition;
import java.util.List;
import eu.dnetlib.pace.config.Cond;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.model.FieldDef;
@ -11,11 +10,12 @@ import eu.dnetlib.pace.model.FieldDef;
*
* @author claudio
*/
@ConditionClass("doiExactMatch")
public class DoiExactMatch extends ExactMatchIgnoreCase {
public final String PREFIX = "(http:\\/\\/dx\\.doi\\.org\\/)|(doi:)";
public DoiExactMatch(final Cond cond, final List<FieldDef> fields) {
public DoiExactMatch(final String cond, final List<FieldDef> fields) {
super(cond, fields);
}

View File

@ -2,7 +2,6 @@ package eu.dnetlib.pace.condition;
import java.util.List;
import eu.dnetlib.pace.config.Cond;
import eu.dnetlib.pace.distance.eval.ConditionEval;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.model.FieldDef;
@ -13,12 +12,17 @@ import org.apache.commons.lang.StringUtils;
*
* @author claudio
*/
@ConditionClass("exactMatch")
public class ExactMatch extends AbstractCondition {
public ExactMatch(final Cond cond, final List<FieldDef> fields) {
public ExactMatch(final String cond, final List<FieldDef> fields) {
super(cond, fields);
}
public ExactMatch(){
super();
}
@Override
protected ConditionEval verify(final FieldDef fd, final Field a, final Field b) {

View File

@ -2,7 +2,6 @@ package eu.dnetlib.pace.condition;
import java.util.List;
import eu.dnetlib.pace.config.Cond;
import eu.dnetlib.pace.distance.eval.ConditionEval;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.model.FieldDef;
@ -12,9 +11,10 @@ import eu.dnetlib.pace.model.FieldDef;
*
* @author claudio
*/
@ConditionClass("exactMatchIgnoreCase")
public class ExactMatchIgnoreCase extends AbstractCondition {
public ExactMatchIgnoreCase(final Cond cond, final List<FieldDef> fields) {
public ExactMatchIgnoreCase(final String cond, final List<FieldDef> fields) {
super(cond, fields);
}

View File

@ -3,7 +3,6 @@ package eu.dnetlib.pace.condition;
import java.util.List;
import com.google.common.collect.Iterables;
import eu.dnetlib.pace.config.Cond;
import eu.dnetlib.pace.distance.eval.ConditionEval;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.model.FieldDef;
@ -13,6 +12,7 @@ import eu.dnetlib.pace.model.FieldDef;
*
* @author claudio
*/
@ConditionClass("mustBeDifferent")
public class MustBeDifferent extends AbstractCondition {
/**
@ -20,7 +20,7 @@ public class MustBeDifferent extends AbstractCondition {
*
* @param fields the fields
*/
public MustBeDifferent(final Cond cond, final List<FieldDef> fields) {
public MustBeDifferent(final String cond, final List<FieldDef> fields) {
super(cond, fields);
}

View File

@ -6,7 +6,6 @@ import java.util.Set;
import java.util.stream.Collectors;
import com.google.common.collect.Sets;
import eu.dnetlib.pace.config.Cond;
import eu.dnetlib.pace.distance.eval.ConditionEval;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.model.FieldDef;
@ -20,11 +19,12 @@ import org.apache.commons.logging.LogFactory;
*
* @author claudio
*/
@ConditionClass("pidMatch")
public class PidMatch extends AbstractCondition {
private static final Log log = LogFactory.getLog(PidMatch.class);
public PidMatch(final Cond cond, final List<FieldDef> fields) {
public PidMatch(final String cond, final List<FieldDef> fields) {
super(cond, fields);
}

View File

@ -4,7 +4,6 @@ import java.util.List;
import com.google.common.collect.Iterables;
import eu.dnetlib.pace.config.Cond;
import eu.dnetlib.pace.distance.eval.ConditionEval;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.model.FieldDef;
@ -14,6 +13,7 @@ import eu.dnetlib.pace.model.FieldDef;
*
* @author claudio
*/
@ConditionClass("sizeMatch")
public class SizeMatch extends AbstractCondition {
/**
@ -22,7 +22,7 @@ public class SizeMatch extends AbstractCondition {
* @param fields
* the fields
*/
public SizeMatch(final Cond cond, final List<FieldDef> fields) {
public SizeMatch(final String cond, final List<FieldDef> fields) {
super(cond, fields);
}

View File

@ -2,7 +2,6 @@ package eu.dnetlib.pace.condition;
import java.util.List;
import eu.dnetlib.pace.config.Cond;
import eu.dnetlib.pace.distance.eval.ConditionEval;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.model.FieldDef;
@ -13,9 +12,10 @@ import eu.dnetlib.pace.model.FieldDef;
* @author claudio
*
*/
@ConditionClass("titleVersionMatch")
public class TitleVersionMatch extends AbstractCondition {
public TitleVersionMatch(final Cond cond, final List<FieldDef> fields) {
public TitleVersionMatch(final String cond, final List<FieldDef> fields) {
super(cond, fields);
}

View File

@ -1,8 +1,8 @@
package eu.dnetlib.pace.condition;
import java.time.Year;
import java.util.List;
import eu.dnetlib.pace.config.Cond;
import eu.dnetlib.pace.distance.eval.ConditionEval;
import org.apache.commons.lang.StringUtils;
@ -14,14 +14,17 @@ import eu.dnetlib.pace.model.FieldDef;
*
* @author claudio
*/
@ConditionClass("yearMatch")
public class YearMatch extends AbstractCondition {
private int limit = 4;
public YearMatch(final Cond cond, final List<FieldDef> fields) {
public YearMatch(final String cond, final List<FieldDef> fields) {
super(cond, fields);
}
public YearMatch(){}
// @Override
// public boolean verify(final Document a, final Document b) {
// boolean res = true;

View File

@ -1,46 +0,0 @@
package eu.dnetlib.pace.config;
/**
* Enumerates the distance Algos.
*/
public enum Algo {
/** The Jaro winkler. */
JaroWinkler,
/** The Jaro winkler title. */
JaroWinklerTitle,
/** The Levenstein. */
Levenstein,
/** The Levenstein distance for title matching */
LevensteinTitle,
/** The Level2 jaro winkler. */
Level2JaroWinkler,
/** The Level2 jaro winkler for title matching */
Level2JaroWinklerTitle,
/** The Level2 levenstein. */
Level2Levenstein,
/** The Sub string levenstein. */
SubStringLevenstein,
/** The Year levenstein. */
YearLevenstein,
/** The Sorted jaro winkler. */
SortedJaroWinkler,
/** The Sorted level2 jaro winkler. */
SortedLevel2JaroWinkler,
/** Compares two urls */
urlMatcher,
/** Exact match algo. */
ExactMatch,
/**
* Returns 0 for equal strings, 1 for different strings.
*/
MustBeDifferent,
/** Always return 1.0 as distance. */
AlwaysMatch,
/** Person distance */
PersonCoAuthorSurnamesDistance,
PersonCoAnchorsDistance,
PersonDistance,
/** The Null. */
Null
}

View File

@ -1,28 +0,0 @@
package eu.dnetlib.pace.config;
/**
* The Enum Cond.
*/
public enum Cond {
/** The year match. */
yearMatch,
/** The title version match. */
titleVersionMatch,
/** The size match. */
sizeMatch,
/**
* Returns true if the field values are different
*/
mustBeDifferent,
/** The Exact match. */
exactMatch,
/**
* The Exact match ignore case.
*/
exactMatchIgnoreCase,
/** The Exact match specialized to recognize DOI values. */
doiExactMatch,
/** The Exact match that checks if pid type and value are the same */
pidMatch
}

View File

@ -2,6 +2,8 @@ package eu.dnetlib.pace.distance;
import eu.dnetlib.pace.model.Field;
import java.util.Map;
/**
* Each field is configured with a distance algo which knows how to compute the distance (0-1) between the fields of two
* objects.
@ -11,5 +13,9 @@ public interface DistanceAlgo {
public abstract double distance(Field a, Field b);
public double getWeight();
public Map<String, Number> getParams();
public void setWeight(double w);
public void setParams(Map<String, Number> params);
}

View File

@ -0,0 +1,13 @@
package eu.dnetlib.pace.distance;
import java.lang.annotation.ElementType;
import java.lang.annotation.Retention;
import java.lang.annotation.RetentionPolicy;
import java.lang.annotation.Target;
@Retention(RetentionPolicy.RUNTIME)
@Target(ElementType.TYPE)
public @interface DistanceClass {
public String value();
}

View File

@ -0,0 +1,24 @@
package eu.dnetlib.pace.distance;
import java.io.Serializable;
import java.lang.reflect.InvocationTargetException;
import java.util.Map;
import java.util.stream.Collectors;
import org.reflections.Reflections;
public class DistanceResolver implements Serializable {
private final Map<String, Class<DistanceAlgo>> functionMap;
public DistanceResolver() {
this.functionMap = new Reflections("eu.dnetlib").getTypesAnnotatedWith(DistanceClass.class).stream()
.filter(DistanceAlgo.class::isAssignableFrom)
.collect(Collectors.toMap(cl -> cl.getAnnotation(DistanceClass.class).value(), cl -> (Class<DistanceAlgo>)cl));
}
public DistanceAlgo resolve(String algo) throws IllegalAccessException, InstantiationException {
return functionMap.get(algo).newInstance();
}
}

View File

@ -25,7 +25,7 @@ public class DistanceScorer {
}
public ScoreResult distance(final Document a, final Document b) {
final ScoreResult sr = new ScoreResult();
final ScoreResult sr = new ScoreResult(); //to keep track of the result of the comparison
sr.setStrictConditions(verify(a, b, config.strictConditions()));
sr.setConditions(verify(a, b, config.conditions()));

View File

@ -1,6 +1,8 @@
package eu.dnetlib.pace.distance;
import java.io.Serializable;
import java.util.List;
import java.util.Map;
import com.wcohen.ss.AbstractStringDistance;
@ -24,6 +26,27 @@ public abstract class SecondStringDistanceAlgo extends AbstractPaceFunctions imp
/** The weight. */
protected double weight = 0.0;
private Map<String, Number> params;
protected SecondStringDistanceAlgo(){
}
protected SecondStringDistanceAlgo(Map<String, Number> params){
this.params = params;
}
public void setWeight(double w){
this.weight = w;
}
public Map<String, Number> getParams(){
return this.params;
}
public void setParams(Map<String, Number> params){
this.params = params;
}
/**
* Instantiates a new second string distance algo.
*
@ -37,6 +60,10 @@ public abstract class SecondStringDistanceAlgo extends AbstractPaceFunctions imp
this.weight = weight;
}
protected SecondStringDistanceAlgo(final AbstractStringDistance ssalgo){
this.ssalgo = ssalgo;
}
/**
* Normalize.
*

View File

@ -1,10 +1,22 @@
package eu.dnetlib.pace.distance.algo;
import com.wcohen.ss.AbstractStringDistance;
import eu.dnetlib.pace.distance.DistanceClass;
import eu.dnetlib.pace.distance.SecondStringDistanceAlgo;
import java.util.Map;
@DistanceClass("AlwaysMatch")
public class AlwaysMatch extends SecondStringDistanceAlgo {
public AlwaysMatch(){
super();
}
public AlwaysMatch(final Map<String, Number> params){
super(params);
}
public AlwaysMatch(final double weight) {
super(weight, new com.wcohen.ss.JaroWinkler());
}

View File

@ -1,10 +1,22 @@
package eu.dnetlib.pace.distance.algo;
import com.wcohen.ss.AbstractStringDistance;
import eu.dnetlib.pace.distance.DistanceClass;
import eu.dnetlib.pace.distance.SecondStringDistanceAlgo;
import java.util.Map;
@DistanceClass("ExactMatch")
public class ExactMatch extends SecondStringDistanceAlgo {
public ExactMatch(){
super();
}
public ExactMatch(Map<String, Number> params){
super(params);
}
public ExactMatch(final double weight) {
super(weight, new com.wcohen.ss.JaroWinkler());
}

View File

@ -1,11 +1,23 @@
package eu.dnetlib.pace.distance.algo;
import com.wcohen.ss.AbstractStringDistance;
import eu.dnetlib.pace.distance.DistanceClass;
import eu.dnetlib.pace.distance.SecondStringDistanceAlgo;
import java.util.Map;
//case class JaroWinkler(w: Double) extends SecondStringDistanceAlgo(w, new com.wcohen.ss.JaroWinkler())
@DistanceClass("JaroWinkler")
public class JaroWinkler extends SecondStringDistanceAlgo {
public JaroWinkler(){
super();
}
public JaroWinkler(Map<String, Number> params){
super(params);
}
public JaroWinkler(double weight) {
super(weight, new com.wcohen.ss.JaroWinkler());
}

View File

@ -1,11 +1,23 @@
package eu.dnetlib.pace.distance.algo;
import com.wcohen.ss.AbstractStringDistance;
import eu.dnetlib.pace.distance.DistanceClass;
import eu.dnetlib.pace.distance.SecondStringDistanceAlgo;
import java.util.Map;
//case class JaroWinkler(w: Double) extends SecondStringDistanceAlgo(w, new com.wcohen.ss.JaroWinkler())
@DistanceClass("JaroWinklerTitle")
public class JaroWinklerTitle extends SecondStringDistanceAlgo {
public JaroWinklerTitle(){
super();
}
public JaroWinklerTitle(Map<String, Number> params){
super(params);
}
public JaroWinklerTitle(double weight) {
super(weight, new com.wcohen.ss.JaroWinkler());
}

View File

@ -1,8 +1,10 @@
package eu.dnetlib.pace.distance.algo;
import com.wcohen.ss.AbstractStringDistance;
import eu.dnetlib.pace.distance.DistanceClass;
import eu.dnetlib.pace.distance.SecondStringDistanceAlgo;
@DistanceClass("Level2JaroWinkler")
public class Level2JaroWinkler extends SecondStringDistanceAlgo {
public Level2JaroWinkler(double w) {

View File

@ -1,8 +1,10 @@
package eu.dnetlib.pace.distance.algo;
import com.wcohen.ss.AbstractStringDistance;
import eu.dnetlib.pace.distance.DistanceClass;
import eu.dnetlib.pace.distance.SecondStringDistanceAlgo;
@DistanceClass("Level2JaroWinklerTitle")
public class Level2JaroWinklerTitle extends SecondStringDistanceAlgo {
public Level2JaroWinklerTitle(final double w) {

View File

@ -1,8 +1,10 @@
package eu.dnetlib.pace.distance.algo;
import com.wcohen.ss.AbstractStringDistance;
import eu.dnetlib.pace.distance.DistanceClass;
import eu.dnetlib.pace.distance.SecondStringDistanceAlgo;
@DistanceClass("Level2Levenstein")
public class Level2Levenstein extends SecondStringDistanceAlgo {
public Level2Levenstein(double w) {

View File

@ -1,10 +1,16 @@
package eu.dnetlib.pace.distance.algo;
import com.wcohen.ss.AbstractStringDistance;
import eu.dnetlib.pace.distance.DistanceClass;
import eu.dnetlib.pace.distance.SecondStringDistanceAlgo;
@DistanceClass("Levenstein")
public class Levenstein extends SecondStringDistanceAlgo {
public Levenstein(){
super(new com.wcohen.ss.Levenstein());
}
public Levenstein(double w) {
super(w, new com.wcohen.ss.Levenstein());
}

View File

@ -1,10 +1,16 @@
package eu.dnetlib.pace.distance.algo;
import com.wcohen.ss.AbstractStringDistance;
import eu.dnetlib.pace.distance.DistanceClass;
import eu.dnetlib.pace.distance.SecondStringDistanceAlgo;
@DistanceClass("LevensteinTitle")
public class LevensteinTitle extends SecondStringDistanceAlgo {
public LevensteinTitle(){
super(new com.wcohen.ss.Levenstein());
}
public LevensteinTitle(final double w) {
super(w, new com.wcohen.ss.Levenstein());
}

View File

@ -1,8 +1,10 @@
package eu.dnetlib.pace.distance.algo;
import com.wcohen.ss.AbstractStringDistance;
import eu.dnetlib.pace.distance.DistanceClass;
import eu.dnetlib.pace.distance.SecondStringDistanceAlgo;
@DistanceClass("MustBeDifferent")
public class MustBeDifferent extends SecondStringDistanceAlgo {
public MustBeDifferent(final double weight) {

View File

@ -1,12 +1,16 @@
package eu.dnetlib.pace.distance.algo;
import eu.dnetlib.pace.distance.DistanceAlgo;
import eu.dnetlib.pace.distance.DistanceClass;
import eu.dnetlib.pace.model.Field;
import java.util.Map;
/**
* Not all fields of a document need to partecipate in the distance measure. We model those fields as having a
* NullDistanceAlgo.
*/
@DistanceClass("Null")
public class NullDistanceAlgo implements DistanceAlgo {
@Override
@ -19,4 +23,16 @@ public class NullDistanceAlgo implements DistanceAlgo {
return 0.0;
}
@Override
public void setWeight(double w){
}
@Override
public Map<String, Number> getParams() {
return null;
}
@Override
public void setParams(Map<String, Number> params) {
}
}

View File

@ -1,10 +1,12 @@
package eu.dnetlib.pace.distance.algo;
import com.wcohen.ss.AbstractStringDistance;
import eu.dnetlib.pace.distance.DistanceClass;
/**
* The Class SortedJaroWinkler.
*/
@DistanceClass("SortedJaroWinkler")
public class SortedJaroWinkler extends SortedSecondStringDistanceAlgo {
/**

View File

@ -1,10 +1,12 @@
package eu.dnetlib.pace.distance.algo;
import com.wcohen.ss.AbstractStringDistance;
import eu.dnetlib.pace.distance.DistanceClass;
/**
* The Class SortedJaroWinkler.
*/
@DistanceClass("Sorted2JaroWinkler")
public class SortedLevel2JaroWinkler extends SortedSecondStringDistanceAlgo {
/**

View File

@ -1,5 +1,6 @@
package eu.dnetlib.pace.distance.algo;
import eu.dnetlib.pace.distance.DistanceClass;
import eu.dnetlib.pace.distance.SecondStringDistanceAlgo;
import org.apache.commons.lang.StringUtils;
@ -8,14 +9,21 @@ import com.wcohen.ss.AbstractStringDistance;
import eu.dnetlib.pace.config.Type;
import eu.dnetlib.pace.model.Field;
import java.util.Map;
/**
* The Class SubStringLevenstein.
*/
@DistanceClass("SubStringLevenstein")
public class SubStringLevenstein extends SecondStringDistanceAlgo {
/** The limit. */
protected int limit;
public SubStringLevenstein() {
super(new com.wcohen.ss.Levenstein());
}
/**
* Instantiates a new sub string levenstein.
*
@ -87,4 +95,9 @@ public class SubStringLevenstein extends SecondStringDistanceAlgo {
return 1 / Math.pow(Math.abs(d) + 1, 0.1);
}
public void setParams(Map<String, Number> params){
this.limit = params.get("limit").intValue(); //necessary because this class needs also the limit
super.setParams(params);
}
}

View File

@ -1,5 +1,6 @@
package eu.dnetlib.pace.distance.algo;
import eu.dnetlib.pace.distance.DistanceClass;
import eu.dnetlib.pace.model.Field;
import org.apache.commons.lang.StringUtils;
@ -7,15 +8,24 @@ import java.net.MalformedURLException;
import java.net.URL;
import java.util.Map;
@DistanceClass("urlMatcher")
public class UrlMatcher extends Levenstein {
private Map<String, Number> params;
public UrlMatcher(){
super();
}
public UrlMatcher(double weight, Map<String, Number> params) {
super(weight);
this.params = params;
}
public void setParams(Map<String, Number> params) {
this.params = params;
}
@Override
public double distance(Field a, Field b) {

View File

@ -1,6 +1,5 @@
package eu.dnetlib.pace.distance.eval;
import eu.dnetlib.pace.config.Cond;
import eu.dnetlib.pace.model.Field;
/**
@ -8,7 +7,7 @@ import eu.dnetlib.pace.model.Field;
*/
public class ConditionEval {
private Cond cond;
private String cond;
private Field a;
@ -16,7 +15,7 @@ public class ConditionEval {
private int result;
public ConditionEval(final Cond cond, final Field a, final Field b, final int result) {
public ConditionEval(final String cond, final Field a, final Field b, final int result) {
this.cond = cond;
this.a = a;
this.b = b;
@ -47,11 +46,11 @@ public class ConditionEval {
this.result = result;
}
public Cond getCond() {
public String getCond() {
return cond;
}
public void setCond(final Cond cond) {
public void setCond(final String cond) {
this.cond = cond;
}
}

View File

@ -1,6 +1,5 @@
package eu.dnetlib.pace.distance.eval;
import eu.dnetlib.pace.config.Algo;
import eu.dnetlib.pace.model.Field;
import eu.dnetlib.pace.model.FieldDef;

View File

@ -2,10 +2,12 @@ package eu.dnetlib.pace.distance.eval;
import com.google.gson.GsonBuilder;
import java.io.Serializable;
/**
* Created by claudio on 09/03/16.
*/
public class ScoreResult {
public class ScoreResult implements Serializable {
private ConditionEvalMap strictConditions;
@ -49,8 +51,12 @@ public class ScoreResult {
@Override
public String toString() {
final GsonBuilder b = new GsonBuilder();
b.serializeSpecialFloatingPointValues();
return b.setPrettyPrinting().create().toJson(this);
//TODO cannot print: why?
// final GsonBuilder b = new GsonBuilder()
// .serializeSpecialFloatingPointValues()
// .serializeNulls();
//
// return b.setPrettyPrinting().create().toJson(this);
return "{}";
}
}

View File

@ -1,6 +1,7 @@
package eu.dnetlib.pace.model;
import java.io.Serializable;
import java.lang.reflect.InvocationTargetException;
import java.util.List;
import java.util.Map;
@ -9,49 +10,36 @@ import eu.dnetlib.pace.clustering.*;
public class ClusteringDef implements Serializable {
private Clustering name;
private String name;
private List<String> fields;
private Map<String, Integer> params;
private ClusteringResolver clusteringResolver = new ClusteringResolver();
public ClusteringDef() {}
public Clustering getName() {
public String getName() {
return name;
}
public void setName(final Clustering name) {
public void setName(final String name) {
this.name = name;
}
public ClusteringFunction getClusteringFunction() {
switch (getName()) {
case acronyms:
return new Acronyms(getParams());
case ngrams:
return new Ngrams(getParams());
case ngrampairs:
return new NgramPairs(getParams());
case sortedngrampairs:
return new SortedNgramPairs(getParams());
case suffixprefix:
return new SuffixPrefix(getParams());
case spacetrimmingfieldvalue:
return new SpaceTrimmingFieldValue(getParams());
case immutablefieldvalue:
return new ImmutableFieldValue(getParams());
case personhash:
return new PersonHash(getParams());
case personclustering:
return new PersonClustering(getParams());
case lowercase:
return new LowercaseClustering(getParams());
case urlclustering:
return new UrlClustering(getParams());
default:
try {
ClusteringFunction clusteringFunction = clusteringResolver.resolve(getName());
clusteringFunction.setParams(params);
return clusteringFunction;
} catch (IllegalAccessException | InstantiationException | NoSuchMethodException | InvocationTargetException e) {
e.printStackTrace();
return new RandomClusteringFunction(getParams());
}
}
public List<String> getFields() {

View File

@ -5,44 +5,36 @@ import java.util.List;
import com.google.gson.Gson;
import eu.dnetlib.pace.condition.*;
import eu.dnetlib.pace.config.Cond;
public class CondDef implements Serializable {
private Cond name;
private String name;
private List<String> fields;
private ConditionResolver conditionResolver = new ConditionResolver();
public CondDef() {}
public ConditionAlgo getConditionAlgo(final List<FieldDef> fields) {
switch (getName()) {
case yearMatch:
return new YearMatch(getName(), fields);
case titleVersionMatch:
return new TitleVersionMatch(getName(), fields);
case sizeMatch:
return new SizeMatch(getName(), fields);
case exactMatch:
return new ExactMatch(getName(), fields);
case mustBeDifferent:
return new MustBeDifferent(getName(), fields);
case exactMatchIgnoreCase:
return new ExactMatchIgnoreCase(getName(), fields);
case doiExactMatch:
return new DoiExactMatch(getName(), fields);
case pidMatch:
return new PidMatch(getName(), fields);
default:
try {
ConditionAlgo conditionAlgo = conditionResolver.resolve(getName());
conditionAlgo.setFields(fields);
conditionAlgo.setCond(getName());
return conditionAlgo;
} catch (IllegalAccessException | InstantiationException e) {
e.printStackTrace();
return new AlwaysTrueCondition(getName(), fields);
}
}
public Cond getName() {
public String getName() {
return name;
}
public void setName(final Cond name) {
public void setName(final String name) {
this.name = name;
}

View File

@ -1,13 +1,13 @@
package eu.dnetlib.pace.model;
import java.io.Serializable;
import java.util.HashMap;
import java.util.List;
import java.util.Map;
import com.google.common.base.Splitter;
import com.google.common.collect.Lists;
import com.google.gson.Gson;
import eu.dnetlib.pace.config.Algo;
import eu.dnetlib.pace.config.Type;
import eu.dnetlib.pace.distance.*;
import eu.dnetlib.pace.distance.algo.*;
@ -19,7 +19,7 @@ public class FieldDef implements Serializable {
public final static String PATH_SEPARATOR = "/";
private Algo algo;
private String algo;
private String name;
@ -37,6 +37,8 @@ public class FieldDef implements Serializable {
private Map<String, Number> params;
private DistanceResolver distanceResolver = new DistanceResolver();
public FieldDef() {}
// def apply(s: String): Field[A]
@ -66,40 +68,22 @@ public class FieldDef implements Serializable {
}
public DistanceAlgo getDistanceAlgo() {
switch (getAlgo()) {
case JaroWinkler:
return new JaroWinkler(getWeight());
case JaroWinklerTitle:
return new JaroWinklerTitle(getWeight());
case Level2JaroWinkler:
return new Level2JaroWinkler(getWeight());
case Level2JaroWinklerTitle:
return new Level2JaroWinklerTitle(getWeight());
case Level2Levenstein:
return new Level2Levenstein(getWeight());
case Levenstein:
return new Levenstein(getWeight());
case LevensteinTitle:
return new LevensteinTitle(getWeight());
case SubStringLevenstein:
return new SubStringLevenstein(getWeight(), getLimit());
case SortedJaroWinkler:
return new SortedJaroWinkler(getWeight());
case SortedLevel2JaroWinkler:
return new SortedLevel2JaroWinkler(getWeight());
case urlMatcher:
return new UrlMatcher(getWeight(), getParams());
case ExactMatch:
return new ExactMatch(getWeight());
case MustBeDifferent:
return new MustBeDifferent(getWeight());
case AlwaysMatch:
return new AlwaysMatch(getWeight());
case Null:
return new NullDistanceAlgo();
default:
try {
if (params == null) {
params = new HashMap<>();
}
params.put("limit", getLimit());
params.put("weight", getWeight());
DistanceAlgo distanceAlgo = distanceResolver.resolve(getAlgo());
distanceAlgo.setParams(params);
distanceAlgo.setWeight(getWeight());
return distanceAlgo;
} catch (IllegalAccessException | InstantiationException e) {
e.printStackTrace();
return new NullDistanceAlgo();
}
}
public boolean isIgnoreMissing() {
@ -135,11 +119,11 @@ public class FieldDef implements Serializable {
this.weight = weight;
}
public Algo getAlgo() {
public String getAlgo() {
return algo;
}
public void setAlgo(final Algo algo) {
public void setAlgo(final String algo) {
this.algo = algo;
}

View File

@ -23,7 +23,6 @@ public class BlockProcessor {
private DedupConfig dedupConf;
public static void constructAccumulator( final DedupConfig dedupConf) {
accumulators.add(String.format("%s::%s",dedupConf.getWf().getEntityType(), "records per hash key = 1"));
accumulators.add(String.format("%s::%s",dedupConf.getWf().getEntityType(), "missing " + dedupConf.getWf().getOrderField()));

Binary file not shown.