antonis.lempesis
/
dnet-hadoop
forked from D-Net/dnet-hadoop
1
1
Fork 1
Code Issues Pull Requests Projects Releases Wiki Activity

code formatting

This commit is contained in:
Claudio Atzori 2020-05-23 09:51:48 +02:00
parent 5d7ac78c41
commit 7181807e64
2 changed files with 14 additions and 13 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

25
dhp-workflows/dhp-dedup-openaire/src/main/java/eu/dnetlib/dhp/oa/dedup/AuthorMerger.java
View File

@ -82,10 +82,11 @@ public class AuthorMerger {
.map(ba -> new Tuple2<>(sim(ba, a._2()), ba)) .map(ba -> new Tuple2<>(sim(ba, a._2()), ba))
.max(Comparator.comparing(Tuple2::_1)); .max(Comparator.comparing(Tuple2::_1));
if(simAuthor.isPresent()) { if (simAuthor.isPresent()) {
double th = THRESHOLD; double th = THRESHOLD;
//increase the threshold if the surname is too short // increase the threshold if the surname is too short
if (simAuthor.get()._2().getSurname() != null && simAuthor.get()._2().getSurname().length()<=3) if (simAuthor.get()._2().getSurname() != null
&& simAuthor.get()._2().getSurname().length() <= 3)
th = 0.99; th = 0.99;
if (simAuthor.get()._1() > th) { if (simAuthor.get()._1() > th) {
@ -100,9 +101,10 @@ public class AuthorMerger {
} }
public static String pidToComparableString(StructuredProperty pid) { public static String pidToComparableString(StructuredProperty pid) {
return (pid.getQualifier() != null ? return (pid.getQualifier() != null
pid.getQualifier().getClassid() != null ? pid.getQualifier().getClassid().toLowerCase() : "" : "") ? pid.getQualifier().getClassid() != null ? pid.getQualifier().getClassid().toLowerCase() : ""
+ (pid.getValue() != null ? pid.getValue().toLowerCase() : ""); : "")
+ (pid.getValue() != null ? pid.getValue().toLowerCase() : "");
} }
public static int countAuthorsPids(List<Author> authors) { public static int countAuthorsPids(List<Author> authors) {
@ -123,14 +125,13 @@ public class AuthorMerger {
final Person pa = parse(a); final Person pa = parse(a);
final Person pb = parse(b); final Person pb = parse(b);
//if both are accurate (e.g. they have name and surname) // if both are accurate (e.g. they have name and surname)
if (pa.isAccurate() & pb.isAccurate()) { if (pa.isAccurate() & pb.isAccurate()) {
return return new JaroWinkler().score(normalize(pa.getSurnameString()), normalize(pb.getSurnameString())) * 0.5
new JaroWinkler().score(normalize(pa.getSurnameString()), normalize(pb.getSurnameString()))*0.5 + new JaroWinkler().score(normalize(pa.getNameString()), normalize(pb.getNameString())) * 0.5;
+ new JaroWinkler().score(normalize(pa.getNameString()), normalize(pb.getNameString()))*0.5;
} else { } else {
return return new JaroWinkler()
new JaroWinkler().score(normalize(pa.getNormalisedFullname()), normalize(pb.getNormalisedFullname())); .score(normalize(pa.getNormalisedFullname()), normalize(pb.getNormalisedFullname()));
} }
} }

2
dhp-workflows/dhp-dedup-openaire/src/test/java/eu/dnetlib/dhp/oa/dedup/EntityMergerTest.java
View File

@ -96,7 +96,7 @@ public class EntityMergerTest implements Serializable {
public void publicationMergerTest2() throws InstantiationException, IllegalAccessException, IOException { public void publicationMergerTest2() throws InstantiationException, IllegalAccessException, IOException {
Publication pub_merged = DedupRecordFactory Publication pub_merged = DedupRecordFactory
.entityMerger(dedupId, publications2.iterator(), 0, dataInfo, Publication.class); .entityMerger(dedupId, publications2.iterator(), 0, dataInfo, Publication.class);
assertEquals(pub_merged.getAuthor().size(), 27); assertEquals(pub_merged.getAuthor().size(), 27);
// insert assertions here // insert assertions here