wnxml.dtd

<!ELEMENT WNXML (SYNSET+)>

<!ELEMENT SYNSET (ID, ID3?, POS, SYNONYM, ILR*, DEF?, BCS?, USAGE*, SNOTE*, STAMP?, DOMAIN?, SUMO*, NL?, TNL?, ELR*, ELR3*, EQ_NEAR_SYNONYM*, EQ_HYPERNYM*, EQ_HYPONYM*, EKSZ*, VFRAME*)>

<!ELEMENT ID (#PCDATA)>

<!-- Synset ID mapped to Princeton WordNet 3.0 -->
<!ELEMENT ID3 (#PCDATA)>

<!ELEMENT POS (#PCDATA)>

<!ELEMENT SYNONYM (LITERAL*)>
<!ELEMENT LITERAL (#PCDATA | SENSE | LNOTE | NUCLEUS)*>
<!ELEMENT SENSE (#PCDATA)>
<!ELEMENT LNOTE (#PCDATA)>
<!ELEMENT NUCLEUS (#PCDATA)>

<!ELEMENT ILR (#PCDATA | TYPE)*>
<!ELEMENT TYPE (#PCDATA)>

<!ELEMENT DEF (#PCDATA)>
<!ELEMENT BCS (#PCDATA)>
<!ELEMENT USAGE (#PCDATA)>
<!ELEMENT SNOTE (#PCDATA)>
<!ELEMENT STAMP (#PCDATA)>
<!ELEMENT DOMAIN (#PCDATA)>
<!ELEMENT SUMO (#PCDATA | TYPE)*>
<!ELEMENT NL (#PCDATA)>

<!-- "Technical Non-Lexicalized" -->
<!ELEMENT TNL (#PCDATA)>

<!-- External linguistic relations; pointers to ILI synsets with their types (Princeton WordNet 2.0) -->
<!ELEMENT ELR (#PCDATA | TYPE)*>

<!-- External linguistic relations; pointers to ILI synsets with their types (Princeton WordNet 3.0) -->
<!ELEMENT ELR3 (#PCDATA | TYPE)*>

<!-- these 3 are retained for backward compatibility, ELR tag should be used instead -->
<!ELEMENT EQ_NEAR_SYNONYM (#PCDATA)>
<!ELEMENT EQ_HYPERNYM (#PCDATA)>
<!ELEMENT EQ_HYPONYM (#PCDATA)>

<!-- EKSZ dictinary sense id + pointer type -->
<!ELEMENT EKSZ (#PCDATA | TYPE)*>

<!-- verb frame id + pointer type -->
<!ELEMENT VFRAME (#PCDATA | TYPE)*>