G.E.R.L. logo Contents

home

G.E.R.L. Tool


G.E.R.L. Format

<!-- The original dtd was simplified to the needs of GERL - this means that several attributes and tags were eliminated -->
<!-- Also there were added some tags in order to make the correspondances between Romanian and other languages -->
<!-- Modified ? in " " at Parole-->

<!DOCTYPE LesParole [

<!ELEMENT LesParole - O ( Parole+ , ParoleMultilingue+)

<!ELEMENT Parole - O
( ParoleMorpho, ParoleSyntaxe, ParoleSemant)>
<!ATTLIST Parole
lexiconname CDATA #REQUIRED
language CDATA #REQUIRED
version CDATA #IMPLIED
creationdate1 CDATA #IMPLIED
modificationdate CDATA #IMPLIED
copyright CDATA #IMPLIED>

<!-- ********************************************************* -->
<!-- ***** MORPHOLOGICAL INFORMATION ***** -->
<!-- ***** ParoleMorpho ***** -->
<!-- ********************************************************* -->

<!ELEMENT ParoleMorpho - O
((MuS|MuC|MuAff)* &
GInP* &
CombMF*)>

<!-- ***************************************************  -->
<!-- ******* DEFINITION OF MORPHOLOGICAL UNITS ****** -->
<!-- *************************************************** -->
<!-- covered noun, verb, pronoun, adjective, article, adverbs, numerals, prepositions, conjuctions, particles-->
<!--some of the types (noun, verb, person, numeral), main/subord for the conjuctions-->
<!ELEMENT MuS - O (Entry,Gmu+ & Derivation*)>
<!--The ADPOSITION was tranformed in PREPOSITION and VERBPARTICLE-->
<!ATTLIST MuS
id ID #REQUIRED
gramcat (WITHOUT|NOUN|VERB|
ADJECTIVE|PRONOUN|
ADVERB|PREPOSITION|VERBPARTICLE|
CONJUNCTION|NUMERAL|
ARTICLE) WITHOUT
gramsubcat (WITHOUT|PROPER|
COMMON|MAIN|AUX|MODAL|
COPULA|
POSSESSIVE|DEMONSTRATIVE|
INTERROGATIVE|RELATIVE|RECIPROCAL
REFLEXIVE|PERSONAL|UNDEFINED|NEGATIVE
COORDINATIVE|SUBORDINATIVE|
CARDINAL|ORDINAL|
FRACTIONAL|REPETATIVE|MULTIPLICATIVE|VARIATIVE|
DEFINITE|INDEFINITE|OTHER) WITHOUT
synulist IDREFS #IMPLIED
foreign (YES|NO|NOSPEC) NOSPEC
semulist IDREFS #IMPLIED>
<!--added semulist as attribute. Connection Mu-SemU, no SynU-SemU
-->

<!ELEMENT MuC - O (Entry, RCompos+)>
<!--The ADPOSITION was tranformed in PREPOSITION and VERBPARTICLE-->
<!ATTLIST MuC
id ID #REQUIRED
gramcat (WITHOUT|NOUN|VERB|
ADJECTIVE|PRONOUN|
ADVERB|PREPOSITION|VERBPARTICLE|
CONJUNCTION|NUMERAL|
ARTICLE) WITHOUT
gramsubcat (WITHOUT|PROPER|
COMMON|MAIN|AUX|MODAL|
COPULA|
POSSESSIVE|DEMONSTRATIVE|
INTERROGATIVE|RELATIVE|RECIPROCAL
REFLEXIVE|PERSONAL|UNDEFINED|NEGATIVE
COORDINATIVE|SUBORDINATIVE|
CARDINAL|ORDINAL|
FRACTIONAL|REPETATIVE|MULTIPLICATIVE|VARIATIVE|
DEFINITE|INDEFINITE|OTHER) WITHOUT
synulist IDREFS #IMPLIED
foreign (YES|NO|NOSPEC) NOSPEC
mainword IDREF #IMPLIED
semulist IDREFS #IMPLIED>
<!--added semulist as attribute. Connection Mu-SemU, no SynU-SemU
-->

<!-- A Compound Morphological Unit has no Gmu of its own:
these graphic forms are deduced from the Units
which make up the Compound Unit.
Each Component that participates in the MuC is indicated
by an RCompos relationship.
A MuC consists at least of 2 Rcompos (which the DTD does not show) -->
<!-- foreign attribute is added in both cases (MuS, MuC) -->

<!ELEMENT MuAff  - O (Entry)>
<!ATTLIST MuAff
id ID #REQUIRED
typaff (WITHOUT|PREFIX|
SUFFIX|BASE) WITHOUT>
<!-- The attribute, typaff records the type of
a Morphological Affix Unit; in the case in which an affix
may be typed only within its derivation context, this
attribute will have the value, WITHOUTS.-->

<!ELEMENT Entry O O (#PCDATA)>
<!-- This element is added so that the search is done more rapid, espacially for machine translations -->

<!-- ********************************************** -->
<!-- ******* MORPHOLOGICAL COMPOSITION ********** -->
<!-- ********************************************** -->
<!ELEMENT RCompos - O	EMPTY>
<!ATTLIST RCompos
linearorder NUMBER #REQUIRED
gsepar (ATTAQUEG|HYPHEN|
APOSTROPHE|SPACE|
JOIN|HYPHENSPACE|
HYPHENJOIN|HYPHENAPOSTROPHE|
HYPHENSPACEJOIN|
APOSTROPHEJOIN|
SPACEJOIN) ATTAQUEG
mu IDREF #REQUIRED>

<!-- The attribute 'mu' indicates a MuS/Cont/Agg/C (Um sub-
classes) component which participates in the composition.
The attribute 'linearorder' specifies
the position of the component in the composition.
The attributs 'g/psepar' ("graphic/phonemic seperators"),
gives the list of possible separators which may appear
before the component. -->

<!-- ************************************************** -->
<!-- ********* GRAPHICAL FORM ******* -->
<!-- ************************************************** -->
<!-- Modification: not used anymore -->
<!-- ************************************************** -->
<!-- ***** GRAPHIC SYSTEM OF INFLECTION *** -->
<!-- ************************************************** -->

<!ELEMENT Gmu - O (EMPTY)>
<!ATTLIST Gmu
inp IDREF #REQUIRED>


<!ELEMENT GInP - O (CombMFCif+)>
<!ATTLIST GInP
id ID #REQUIRED
comment CDATA #IMPLIED
example CDATA #IMPLIED

<!ELEMENT CombMFCif - O EMPTY>
<!ATTLIST CombMFCif
combmf IDREF #REQUIRED>

<!-- A CombMFCif refers to a CombTM (Combination of Morphological Features) via the
'combmf' feature.
Added of degree - possible not necessary -->

<!ELEMENT CombMF - O EMPTY>
<!ATTLIST CombMF
id ID #REQUIRED
gender (WITHOUT|MASCULINE|FEMININE|
NEUTER) WITHOUTG
number (WITHOUT|SINGULAR|PLURAL) WITHOUT
case (WITHOUT|NOMINATIVE|GENITIVE|
DATIVE|ACCUSATIVE|VOCATIVE) WITHOUT
mood (WITHOUT|INDICATIVE|IMPERATIVE|
INFINITIVE|PARTICIPLE|GERUND|
CONJUNCTIVE) WITHOUT
tense (WITHOUT|PRESENT|IMPERFECT|
PAST|PLUSQUEPARFAIT|PERFECTSIMPLE) WITHOUT
person (WITHOUT|1|2|3) WITHOUT
reflexivity (WITHOUT|RREFL|NOREFL) WITHOUT
degree (WITHOUT|POSITIVE|
COMPARATIVE|SUPERLATIVE) WITHOUT
degreetype (WITHOUT|SUPERIORITY|INFERIORITY|EQUALITY|ABSOLUTE) WITHOUT
transitivity (WITHOUT|TRANSITIVE|INTRANSITIVE) WITHOUT
typepers (WITHOUT|PERSONAL|IMPERSONAL) WITHOUT
article (WITHOUT|DEFINITE|INDEFINITE) WITHOUT
hasparticle (YES|NO) NO
>
<!--it is covered gender, number, case, mode, time, voice, person, degree, reflxivity, transitivity, (im)personal-->

<!-- ************************************************ -->
<!--	*******  MORPHOLOGICAL DERIVATION    *********** -->
<!-- ************************************************ -->

<!ELEMENT Derivation - O (RDeriv+)>
<!ATTLIST Derivation
comment CDATA #IMPLIED>
<!-- The content token 'RDeriv' is used to record the
different components of a derivation. Concurrent
derivations are indicated by recording several Derivation
elements on one derived Unit.-->

<!ELEMENT RDeriv - O EMPTY>
<!ATTLIST RDeriv
linearorder NUMBER #IMPLIED
status (WITHOUT|PREFIX|
SUFFIX|BASE) WITHOUT
mu IDREF #REQUIRED>
<!-- The field 'mu' indicates the component of the derivation.
The attribute 'linearorder' indicates the range of the Mu in the
derivation -->
<!-- ********************************************************* -->
<!-- ***** SYNTACTIC INFORMATION ***** -->
<!-- ***** PAROLESYNTAXE ***** -->
<!-- ********************************************************* -->

<!-- It is used for Conjuctions and Verbs. The other have the EMPTY SynU-->

<!ELEMENT ParoleSyntaxe - O (
SynU+ &
Description+ &
Construction*)>
<!-- ********************************************************* -->
<!-- ***** SYNTACTIC UNIT, DESCRIPTION ***** -->
<!-- ********************************************************* -->

<!ELEMENT SynU - O EMPTY>
<!ATTLIST SynU
id ID #REQUIRED
comment CDATA #IMPLIED
example CDATA #IMPLIED
description IDREF #REQUIRED>
<!-- SynU describes one syntactic behaviour of a Mu.
One has to encode as many SynU as syntactic behaviours for a same Mu.
- The attribute 'description' records the base description,
- CorrespSynUSemU encodes the correspondence with the semantic level.(CorrespSynUSemU*) from SynU deleted -->

<!ELEMENT Description - O EMPTY>
<!ATTLIST Description
id ID #REQUIRED
comment CDATA #IMPLIED
example CDATA #IMPLIED
representativemu CDATA #IMPLIED
construction IDREF #IMPLIED>
<!-- The attribute 'representativemu' records the id of MU,
the attribute 'construction' records the id of the Construction -->

<!-- ********************************************************* -->
<!-- ***** CONSTRUCTION ***** -->
<!-- ********************************************************* -->

<!ELEMENT Construction - O (SyntFeatureClosed*)>
<!ATTLIST Construction
id ID #REQUIRED
comment CDATA #IMPLIED
example CDATA #IMPLIED>
<!-- A Construction describes the context or syntactic frame specific
to the entry described. -->
<!-- *********************************************************  -->
<!-- ***** FEATURES ***** -->
<!-- ********************************************************* -->

<!ELEMENT SyntFeatureClosed - O EMPTY>
<!ATTLIST SyntFeatureClosed
featurename (CASE|FOLLOWEDBY) #REQUIRED
value (MAIN|SUBORDINATE
NOMINATIVE|GENITIVE|DATIVE|
ACCUSATIVE|VOCATIVE) #REQUIRED >


<!-- ********************************************************* -->
<!-- ***** SEMANTIC INFORMATION ***** -->
<!-- *****                   PAROLESEMANT                ***** -->
<!-- ********************************************************* -->

<!-- It is used for thematic roles - VERBS, synonymy and collocations -->
<!-- Diffrent way of expressing collocations: new tag Collocation!!! -->

<!ELEMENT ParoleSemant - O (

SemU+
& Predicate*
& Argument*
& SemanticRole*
& RSemU*
& Collocation*)>
<!ELEMENT SemU  - O (PredicativeRepresentation?,RWeightValSemU*) >
<!ATTLIST SemU
id ID #REQUIRED
example CDATA #IMPLIED
comment CDATA #IMPLIED
collocationlist IDREFS #IMPLIED>

<!ELEMENT PredicativeRepresentation - O EMPTY>
<!ATTLIST PredicativeRepresentation
predicate IDREF #REQUIRED>
<!-- Used for semantic/thematic roles -->

<!ELEMENT Predicate - O EMPTY>
<!ATTLIST Predicate
id ID #REQUIRED
example CDATA #IMPLIED
comment CDATA #IMPLIED
argumentl IDREFS #REQUIRED>
<!ELEMENT Argument  - O EMPTY>
<!ATTLIST Argument
id ID #REQUIRED
example CDATA #IMPLIED
comment CDATA #IMPLIED
position1 (NO|BEFORE|AFTER) NO
position2 CDATA #IMPLIED
semanticrolel IDREFS #REQUIRED>
<!-- position1, position2 added for meking more clear -->

<!-- added already in the lexicon -->
<!ELEMENT SemanticRole - O EMPTY>
<!ATTLIST SemanticRole
id ID #REQUIRED
example CDATA #IMPLIED
comment CDATA #IMPLIED
name CDATA #REQUIRED>

<!-- used for synonym-relation, modified target as a list because this way partial synonsmy is also taken into consideration-->
<!ELEMENT RWeightValSemU - O EMPTY>
<!ATTLIST RWeightValSemU
comment CDATA #IMPLIED
targetlist IDREFS #REQUIRED
semr IDREF #REQUIRED>

<!-- there is only one type of relation: synonymy
added as such in lexicon - is the only one-->
<!ELEMENT RSemU - O EMPTY>
<!ATTLIST RSemU
id SYN SYN
comment CDATA #IMPLIED
sstype (SYNONYMY) SYNONYMY>

<!-- new added tag so that entering collocations is easier -->
<!ELEMENT Collocation O O EMPTY>
<!ATTLIST Collocation
id ID #REQUIRED
expression CDATA #IMPLIED
meaning CDATA #IMPLIED
synonymMu IDREF #IMPLIED>
<!-- ********************************************************* -->
<!-- ***** MULTILINGUALITY ***** -->
<!-- ********************************************************* -->

<!ELEMENT ParoleMultilingue - O (CorrespMultColloc* & CorrespMultMU* & CorrespGap*)>

<!ATTLIST ParoleMultilingue
langue1 CDATA #REQUIRED
langue2 CDATA #REQUIRED >

<!-- Added tags, almost totaly changed!!! -->
<!ELEMENT  CorrespMultColloc - O (Referent)>
<!ATTLIST CorrespMultColloc
id ID #REQUIRED
commentaire CDATA #IMPLIED
colloclangue1 IDREF #REQUIRED>

<!ELEMENT Referent - O (EMPTY)>
<!ATTLIST Referent
typereferent (MU|TRANSLATION|COLLOCATION|NOTKNOWN) NOTKNOWN
referentref IDREF #IMPLIED
translation CDATA #IMPLIED>
<!-- if there is no link to another one, here is the text
typereferent was transformed from CDATA
-->


<!ELEMENT CorrespMultMU - O EMPTY>
<!ATTLIST CorrespMultMU
id ID #REQUIRED
commentaire CDATA #IMPLIED
mulangue1 IDREF #REQUIRED
mulangue2 IDREFS #REQUIRED>
<!-- mulangue2 transformed from IDREF to IDREFS: translation 1:n.
translation into collocations are given as translation gaps. Otherwise the dtd should be changed

In the new context there is not needed. Each word has only one CombMF
<!ELEMENT CorrespMultMF - O EMPTY>
<!ATTLIST CorrespMultMF
id ID #REQUIRED
commentaire CDATA #IMPLIED
mflangue1 IDREF #REQUIRED
mflangue2 IDREF #REQUIRED>
-->

<!ELEMENT CorrespGap - O EMPTY>
<!ATTLIST CorrespGap
id ID #REQUIRED
commentaire CDATA #IMPLIED
mu IDREF #REQUIRED
translation CDATA #REQUIRED>

<!-- langueWithoutTranslation CDATA #REQUIRED
langueWithTranslation CDATA #REQUIRED
Not used, because translation is only in one direction
-->
]>