lsmyrnaios
/
dnet-hadoop
forked from D-Net/dnet-hadoop
1
0
Fork 0
Code Pull Requests Activity

Merge pull request 'affroNewModelonBeta' (#494) from affroNewModelonBeta into beta 

Reviewed-on: D-Net/dnet-hadoop#494
This commit is contained in:
Claudio Atzori 2024-10-28 10:48:34 +01:00
parent 1cdcd07a7e 842cc75dae
commit d71df6de19
5 changed files with 50 additions and 37 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

49
dhp-workflows/dhp-aggregation/src/main/java/eu/dnetlib/dhp/actionmanager/bipaffiliations/PrepareAffiliationRelations.java
View File

@ -34,7 +34,7 @@ import eu.dnetlib.dhp.schema.oaf.utils.OafMapperUtils;
import scala.Tuple2;
/**
* Creates action sets for Crossref affiliation relations inferred by BIP!
* Creates action sets for Crossref affiliation relations inferred by OpenAIRE
*/
public class PrepareAffiliationRelations implements Serializable {
@ -104,22 +104,22 @@ public class PrepareAffiliationRelations implements Serializable {
.listKeyValues(OPENAIRE_DATASOURCE_ID, OPENAIRE_DATASOURCE_NAME);
JavaPairRDD<Text, Text> crossrefRelations = prepareAffiliationRelationsNewModel(
spark, crossrefInputPath, collectedfromOpenAIRE);
spark, crossrefInputPath, collectedfromOpenAIRE, BIP_INFERENCE_PROVENANCE + "::crossref");
JavaPairRDD<Text, Text> pubmedRelations = prepareAffiliationRelations(
spark, pubmedInputPath, collectedfromOpenAIRE);
spark, pubmedInputPath, collectedfromOpenAIRE, BIP_INFERENCE_PROVENANCE + "::pubmed");
JavaPairRDD<Text, Text> openAPCRelations = prepareAffiliationRelationsNewModel(
spark, openapcInputPath, collectedfromOpenAIRE);
spark, openapcInputPath, collectedfromOpenAIRE, BIP_INFERENCE_PROVENANCE + "::openapc");
JavaPairRDD<Text, Text> dataciteRelations = prepareAffiliationRelations(
spark, dataciteInputPath, collectedfromOpenAIRE);
JavaPairRDD<Text, Text> dataciteRelations = prepareAffiliationRelationsNewModel(
spark, dataciteInputPath, collectedfromOpenAIRE, BIP_INFERENCE_PROVENANCE + "::datacite");
JavaPairRDD<Text, Text> webCrawlRelations = prepareAffiliationRelations(
spark, webcrawlInputPath, collectedfromOpenAIRE);
JavaPairRDD<Text, Text> webCrawlRelations = prepareAffiliationRelationsNewModel(
spark, webcrawlInputPath, collectedfromOpenAIRE, BIP_INFERENCE_PROVENANCE + "::rawaff");
JavaPairRDD<Text, Text> publisherRelations = prepareAffiliationRelationFromPublisher(
spark, publisherlInputPath, collectedfromOpenAIRE);
JavaPairRDD<Text, Text> publisherRelations = prepareAffiliationRelationFromPublisherNewModel(
spark, publisherlInputPath, collectedfromOpenAIRE, BIP_INFERENCE_PROVENANCE + "::webcrawl");
crossrefRelations
.union(pubmedRelations)
@ -133,7 +133,8 @@ public class PrepareAffiliationRelations implements Serializable {
private static JavaPairRDD<Text, Text> prepareAffiliationRelationFromPublisherNewModel(SparkSession spark,
String inputPath,
List<KeyValue> collectedfrom) {
List<KeyValue> collectedfrom,
String dataprovenance) {
Dataset<Row> df = spark
.read()
@ -142,12 +143,13 @@ public class PrepareAffiliationRelations implements Serializable {
.json(inputPath)
.where("DOI is not null");
return getTextTextJavaPairRDD(collectedfrom, df.selectExpr("DOI", "Organizations as Matchings"));
return getTextTextJavaPairRDDNew(
collectedfrom, df.selectExpr("DOI", "Organizations as Matchings"), dataprovenance);
}
private static JavaPairRDD<Text, Text> prepareAffiliationRelationFromPublisher(SparkSession spark, String inputPath,
List<KeyValue> collectedfrom) {
List<KeyValue> collectedfrom, String dataprovenance) {
Dataset<Row> df = spark
.read()
@ -155,13 +157,14 @@ public class PrepareAffiliationRelations implements Serializable {
.json(inputPath)
.where("DOI is not null");
return getTextTextJavaPairRDD(collectedfrom, df.selectExpr("DOI", "Organizations as Matchings"));
return getTextTextJavaPairRDD(
collectedfrom, df.selectExpr("DOI", "Organizations as Matchings"), dataprovenance);
}
private static <I extends Result> JavaPairRDD<Text, Text> prepareAffiliationRelations(SparkSession spark,
String inputPath,
List<KeyValue> collectedfrom) {
List<KeyValue> collectedfrom, String dataprovenance) {
// load and parse affiliation relations from HDFS
Dataset<Row> df = spark
@ -170,12 +173,12 @@ public class PrepareAffiliationRelations implements Serializable {
.json(inputPath)
.where("DOI is not null");
return getTextTextJavaPairRDD(collectedfrom, df);
return getTextTextJavaPairRDD(collectedfrom, df, dataprovenance);
}
private static <I extends Result> JavaPairRDD<Text, Text> prepareAffiliationRelationsNewModel(SparkSession spark,
String inputPath,
List<KeyValue> collectedfrom) {
List<KeyValue> collectedfrom, String dataprovenance) {
// load and parse affiliation relations from HDFS
Dataset<Row> df = spark
.read()
@ -184,10 +187,11 @@ public class PrepareAffiliationRelations implements Serializable {
.json(inputPath)
.where("DOI is not null");
return getTextTextJavaPairRDDNew(collectedfrom, df);
return getTextTextJavaPairRDDNew(collectedfrom, df, dataprovenance);
}
private static JavaPairRDD<Text, Text> getTextTextJavaPairRDD(List<KeyValue> collectedfrom, Dataset<Row> df) {
private static JavaPairRDD<Text, Text> getTextTextJavaPairRDD(List<KeyValue> collectedfrom, Dataset<Row> df,
String dataprovenance) {
// unroll nested arrays
df = df
.withColumn("matching", functions.explode(new Column("Matchings")))
@ -219,7 +223,7 @@ public class PrepareAffiliationRelations implements Serializable {
DataInfo dataInfo = OafMapperUtils
.dataInfo(
false,
BIP_INFERENCE_PROVENANCE,
dataprovenance,
true,
false,
qualifier,
@ -235,7 +239,8 @@ public class PrepareAffiliationRelations implements Serializable {
new Text(OBJECT_MAPPER.writeValueAsString(aa))));
}
private static JavaPairRDD<Text, Text> getTextTextJavaPairRDDNew(List<KeyValue> collectedfrom, Dataset<Row> df) {
private static JavaPairRDD<Text, Text> getTextTextJavaPairRDDNew(List<KeyValue> collectedfrom, Dataset<Row> df,
String dataprovenance) {
// unroll nested arrays
df = df
.withColumn("matching", functions.explode(new Column("Matchings")))
@ -276,7 +281,7 @@ public class PrepareAffiliationRelations implements Serializable {
DataInfo dataInfo = OafMapperUtils
.dataInfo(
false,
BIP_INFERENCE_PROVENANCE,
dataprovenance,
true,
false,
qualifier,

12
dhp-workflows/dhp-aggregation/src/main/resources/eu/dnetlib/dhp/actionmanager/bipaffiliations/job.properties
View File

@ -31,9 +31,11 @@ spark2SqlQueryExecutionListeners=com.cloudera.spark.lineage.NavigatorQueryListen
# The following is needed as a property of a workflow
oozie.wf.application.path=${oozieTopWfApplicationPath}
crossrefInputPath=/data/bip-affiliations/crossref-data.json
pubmedInputPath=/data/bip-affiliations/pubmed-data.json
openapcInputPath=/data/bip-affiliations/openapc-data.json
dataciteInputPath=/data/bip-affiliations/datacite-data.json
crossrefInputPath=/data/openaire-affiliations/crossref-data.json
pubmedInputPath=/data/openaire-affiliations/pubmed-data-v4.json
openapcInputPath=/data/openaire-affiliations/openapc-data.json
dataciteInputPath=/data/openaire-affiliations/datacite-data.json
webCrawlInputPath=/data/openaire-affiliations/webCrawl
publisherInputPath=/data/openaire-affiliations/publishers
outputPath=/tmp/crossref-affiliations-output-v5
outputPath=/tmp/affRoAS

9
dhp-workflows/dhp-aggregation/src/main/resources/eu/dnetlib/dhp/actionmanager/bipaffiliations/oozie_app/workflow.xml
View File

@ -1,4 +1,4 @@
<workflow-app name="BipAffiliations" xmlns="uri:oozie:workflow:0.5">
<workflow-app name="OpenAIREAffiliations" xmlns="uri:oozie:workflow:0.5">
<parameters>
<property>
@ -21,6 +21,10 @@
<name>webCrawlInputPath</name>
<description>the path where to find the inferred affiliation relations from webCrawl</description>
</property>
<property>
<name>publisherInputPath</name>
<description>the path where to find the inferred affiliation relations from publisher websites</description>
</property>
<property>
<name>outputPath</name>
<description>the path where to store the actionset</description>
@ -99,7 +103,7 @@
<spark xmlns="uri:oozie:spark-action:0.2">
<master>yarn</master>
<mode>cluster</mode>
<name>Produces the atomic action with the inferred by BIP! affiliation relations (from Crossref and Pubmed)</name>
<name>Produces the atomic action with the inferred by OpenAIRE affiliation relations</name>
<class>eu.dnetlib.dhp.actionmanager.bipaffiliations.PrepareAffiliationRelations</class>
<jar>dhp-aggregation-${projectVersion}.jar</jar>
<spark-opts>
@ -117,6 +121,7 @@
<arg>--openapcInputPath</arg><arg>${openapcInputPath}</arg>
<arg>--dataciteInputPath</arg><arg>${dataciteInputPath}</arg>
<arg>--webCrawlInputPath</arg><arg>${webCrawlInputPath}</arg>
<arg>--publisherInputPath</arg><arg>${publisherInputPath}</arg>
<arg>--outputPath</arg><arg>${outputPath}</arg>
</spark>
<ok to="End"/>

16
dhp-workflows/dhp-aggregation/src/test/java/eu/dnetlib/dhp/actionmanager/bipaffiliations/PrepareAffiliationRelationsTest.java
View File

@ -98,9 +98,9 @@ public class PrepareAffiliationRelationsTest {
"-crossrefInputPath", crossrefAffiliationRelationPathNew,
"-pubmedInputPath", crossrefAffiliationRelationPath,
"-openapcInputPath", crossrefAffiliationRelationPathNew,
"-dataciteInputPath", crossrefAffiliationRelationPath,
"-webCrawlInputPath", crossrefAffiliationRelationPath,
"-publisherInputPath", publisherAffiliationRelationOldPath,
"-dataciteInputPath", crossrefAffiliationRelationPathNew,
"-webCrawlInputPath", crossrefAffiliationRelationPathNew,
"-publisherInputPath", publisherAffiliationRelationPath,
"-outputPath", outputPath
});
@ -112,7 +112,7 @@ public class PrepareAffiliationRelationsTest {
.map(aa -> ((Relation) aa.getPayload()));
// count the number of relations
assertEquals(150, tmp.count());// 18 + 24 *3 + 30 * 2 =
assertEquals(162, tmp.count());// 18 + 24 + 30 * 4 =
Dataset<Relation> dataset = spark.createDataset(tmp.rdd(), Encoders.bean(Relation.class));
dataset.createOrReplaceTempView("result");
@ -123,7 +123,7 @@ public class PrepareAffiliationRelationsTest {
// verify that we have equal number of bi-directional relations
Assertions
.assertEquals(
75, execVerification
81, execVerification
.filter(
"relClass='" + ModelConstants.HAS_AUTHOR_INSTITUTION + "'")
.collectAsList()
@ -131,7 +131,7 @@ public class PrepareAffiliationRelationsTest {
Assertions
.assertEquals(
75, execVerification
81, execVerification
.filter(
"relClass='" + ModelConstants.IS_AUTHOR_INSTITUTION_OF + "'")
.collectAsList()
@ -158,7 +158,7 @@ public class PrepareAffiliationRelationsTest {
Assertions
.assertEquals(
2, execVerification.filter("source = '" + publisherid + "' and target = '" + rorId + "'").count());
4, execVerification.filter("source = '" + publisherid + "' and target = '" + rorId + "'").count());
Assertions
.assertEquals(
@ -173,7 +173,7 @@ public class PrepareAffiliationRelationsTest {
Assertions
.assertEquals(
3, execVerification
1, execVerification
.filter(
"source = '" + ID_PREFIX
+ IdentifierFactory

1
dhp-workflows/dhp-graph-mapper/src/test/scala/eu/dnetlib/dhp/enrich/orcid/ORCIDAuthorMatchersTest.scala
View File

@ -31,6 +31,7 @@ class ORCIDAuthorMatchersTest {
assertTrue(matchOrderedTokenAndAbbreviations("孙林 Sun Lin", "Sun Lin"))
// assertTrue(AuthorsMatchRevised.compare("孙林 Sun Lin", "孙林")); // not yet implemented
}
@Test def testDocumentationNames(): Unit = {
assertTrue(matchOrderedTokenAndAbbreviations("James C. A. Miller-Jones", "James Antony Miller-Jones"))
}