added title of section to the extraction

sven-h · Jun 5, 2019 · 6ec9abe · 6ec9abe
1 parent cff322a
commit 6ec9abe
Show file tree

Hide file tree

Showing 3 changed files with 27 additions and 5 deletions.
diff --git a/core/src/main/scala/org/dbpedia/extraction/nif/GeneralNifExtractor.scala b/core/src/main/scala/org/dbpedia/extraction/nif/GeneralNifExtractor.scala
@@ -36,6 +36,7 @@ class GeneralNifExtractor (
   val wikiPageExternalLinkProperty = context.ontology.properties("wikiPageExternalLink")
   val wikiPageInterWikiLinkProperty = context.ontology.properties("wikiPageInterWikiLink")
   val wikiPageInterLanguageLinkProperty = context.ontology.properties("wikiPageInterLanguageLink")
+  val labelProperty = context.ontology.properties("rdfs:label")
 
 
 
@@ -69,7 +70,7 @@ class GeneralNifExtractor (
   private def writeLongAndShortAbstract(section: NifSection, text:String):ArrayBuffer[Quad] = {
     var quads = ArrayBuffer[Quad]()
     if (recordAbstracts && section.id == "abstract" && text.length > 0) {
-      val describingParagraphs = getParagraphsDescribingConcept(section, text)
+      val describingParagraphs = section.paragraphs//getParagraphsDescribingConcept(section, text)
       if(describingParagraphs.size > 0){
         quads += longQuad(wikiPage.uri, text.substring(describingParagraphs.head.begin.getOrElse(0), describingParagraphs.last.end.getOrElse(0)), sourceUrl) //text.substring(section.begin.getOrElse(0), section.end.getOrElse(0)), sourceUrl)
         quads += shortQuad(wikiPage.uri, getShortAbstract(describingParagraphs, text), sourceUrl) // getShortAbstract(section.paragraphs, text), sourceUrl)
@@ -184,6 +185,21 @@ class GeneralNifExtractor (
     if (section.next.isEmpty)
       quads += nifStructure(topSectionUri, RdfNamespace.NIF.append("lastSection"), sectionUri, sourceUrl, null)
 
+    //adding title
+    if(section.beginTitle.nonEmpty && section.endTitle.nonEmpty){
+      val titleUri = getNifIri("title", section.beginTitle.get, section.endTitle.get)
+      quads += nifStructure(titleUri, RdfNamespace.RDF.append("type"), RdfNamespace.NIF.append("Title"), sourceUrl, null)
+      quads += nifStructure(titleUri, RdfNamespace.NIF.append("referenceContext"), nifContextUri, sourceUrl, null)
+      quads += nifStructure(titleUri, RdfNamespace.NIF.append("beginIndex"), section.beginTitle.get.toString, sourceUrl, RdfNamespace.XSD.append("nonNegativeInteger"))
+      quads += nifStructure(titleUri, RdfNamespace.NIF.append("endIndex"), section.endTitle.get.toString, sourceUrl, RdfNamespace.XSD.append("nonNegativeInteger"))
+      quads += nifStructure(titleUri, RdfNamespace.NIF.append("superString"), sectionUri, sourceUrl, null)
+      if(writeLinkAnchors){
+        quads += nifStructure(titleUri, RdfNamespace.NIF.append("anchorOf"), section.id, sourceUrl, RdfNamespace.XSD.append("string"))
+        quads += nifStructure(sectionUri, labelProperty.uri, section.id.trim, sourceUrl, RdfNamespace.XSD.append("string"))
+      }
+
+    }
+
     quads
   }
 

diff --git a/core/src/main/scala/org/dbpedia/extraction/nif/NifExtractionAstVisitor.scala b/core/src/main/scala/org/dbpedia/extraction/nif/NifExtractionAstVisitor.scala
@@ -17,7 +17,7 @@ class NifExtractionAstVisitor(language : Language)
   private var currentSection = new ListBuffer[Int]()
 
   private var context = StringBuilder.newBuilder //contains the whole text of the wikipage
-  private var nifSection: NifSection = new NifSection(id = "abstract",ref = "", prev = None, next = None, top = None, sub = None, begin = Some(0), end = None, paragraphs = ListBuffer())
+  private var nifSection: NifSection = new NifSection(id = "abstract",ref = "", prev = None, next = None, top = None, sub = None, begin = Some(0), end = None, beginTitle = None, endTitle = None, paragraphs = ListBuffer())
   private var nifParagraph: NifParagraph = new NifParagraph(begin = Some(0), end = None,links = ListBuffer())
   private var extLinkNum :Int = 1
 
@@ -42,6 +42,8 @@ class NifExtractionAstVisitor(language : Language)
       sub = None,
       begin = Some(0),
       end = None,
+      beginTitle = None,
+      endTitle = None,
       paragraphs = ListBuffer()
     )
     nifSection = abstractSection
@@ -256,6 +258,8 @@ class NifExtractionAstVisitor(language : Language)
       sub = None,
       begin = None,
       end = None,
+      beginTitle = None,
+      endTitle = None,
       paragraphs = ListBuffer()
     )
     section.top match{
@@ -282,10 +286,10 @@ class NifExtractionAstVisitor(language : Language)
       tocMap.append(nifSection)
       //closeParagraphAndStartNew()
       nifParagraph = new NifParagraph(begin = Some(0),end = Some(0),links = ListBuffer()) //dummy paragraph (all lnks in header will be put in dummy)
-      val startHeading = context.length
+      nifSection.beginTitle = Some(context.length)
       iterate(s.getHeading)
-      val endHeading = context.length
-      nifSection.id = context.substring(startHeading, endHeading)
+      nifSection.endTitle = Some(context.length)
+      nifSection.id = context.substring(nifSection.beginTitle.get, nifSection.endTitle.get)
 
       closeParagraphAndStartNew()
     }

diff --git a/core/src/main/scala/org/dbpedia/extraction/nif/NifSection.scala b/core/src/main/scala/org/dbpedia/extraction/nif/NifSection.scala
@@ -11,6 +11,8 @@ class NifSection(
                   var sub: Option[NifSection],
                   var begin: Option[Int],
                   var end: Option[Int],
+                  var beginTitle: Option[Int],
+                  var endTitle: Option[Int],
                   var paragraphs: ListBuffer[NifParagraph]
                 ) {
   def addParagraph(nifparagraph: NifParagraph): Unit = paragraphs += nifparagraph