Corpus Format

Song lyrics in the Songkorpus are encoded in XML TEI P5 format, validated against the Songkorpus DTD. Linguistic metadata — part of speech (extended STTS), lemma, and named entities — is annotated separately using the WebLicht TCF format and the UIMA CAS XMI format.

TEI P5 Header

Each file begins with a <teiHeader> containing bibliographic metadata: title, author(s), publisher, publication date, album, and a link to the artist's source. Additional metadata, such as chart rank, musical genre, or artist gender information, may also be present. 

<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE TEI SYSTEM "https://songkorpus.de/songkorpus.dtd">
<TEI xmlns="http://www.tei-c.org/ns/1.0" version="3.4.0">
  <teiHeader>
    <fileDesc>
      <titleStmt xml:id="SK/STO.2023.5">
        <title>Kommt mal alle wieder runter</title>
        <author role="primary">Stoppok</author>
        <author role="text">Stefan Stoppok</author>
      </titleStmt>
      <publicationStmt>
        <publisher>Grundvermögen Edition</publisher>
        <date>2023</date>
        <ref><name>Teufelsküche</name></ref>
      </publicationStmt>
      <sourceDesc>
        <ab>
          <link target="https://stoppok.de/alben/27/"/>
        </ab>
      </sourceDesc>
    </fileDesc>
  </teiHeader>

TEI P5 Body

The song text is structured using <div1 type="song">, with strophes as <lg type="verse"> and individual lines as <l>. Additional structural types (chorus, bridge, etc.) and spelling annotations may also appear. 

<text>
  <body>
    <div1 type="song">
      <lg type="verse">
        <l>Kommt mal alle wieder runter,</l>
        <l>da oben ist sowieso Ende.</l>
        <l>Es wird jetzt auch nicht mehr bunter hier,</l>
        <l>eher im Gegenteil,</l>
        <l>jetzt kommt erst die eigentliche Wende.</l>
      </lg>
      <lg type="verse">
        <l>Die Frage ist: wer schlägt wem den Schädel ein?</l>
        <l>Wer schleppt noch mehr Kohle heim?</l>
        <l>Wer schreit am lautesten, wer hat den längsten,</l>
        <l>wer hat am wenigsten Skrupel und Bedenken?</l>
      </lg>
    </div1>
  </body>
</text>
</TEI>

Linguistic Annotation Layers

Annotation Type Tool/Format Coverage
Part of speech TreeTagger (extended STTS) All archives
Lemmatization TreeTagger All archives
Named entities German NER / manual annotation All archives
Neologisms manual annotation All artist archives
Constituent structures Tiger Treebank Tagset All artist archives
Rhyme types manual annotation Selected archives