Annotating Adjectives, Adverbs, and Relative Clauses
Predicative adjectives are annotated much like verbs:
Dependency graph for sentence: I am despicable. Token despicable is labeled PROPERTY and has an edge labeled has-property to token I.
Dependency graph for sentence: The dog is tired. Token tired is labeled INTERNAL-STATE and has an edge labeled has-state to token dog.
Dependency graph for sentence: Change is necessary. Token necessary is labeled NECESSITY and has an edge labeled necessary to token Change.
Dependency graph for sentence: Change is possible. Token possible is labeled POSSIBILITY and has an edge labeled possible to token Change.
Attributive adjectives and verbs behave syntactically like modifiers, but they
are also predicates assigning the modifiee a role. For the modification
relation, we use the META frame, assigning the modifying predicate the
scene role and – implicitly – the modifiee the participant role. We then
introduce an additional semantic dependency (we call it a backlink) from the
modifying predicate to the modifiee to annotate the role that the former
assigns to the latter. For example:
Dependency graph for sentence: despicable me. Token despicable is labeled PROPERTY and has an edge labeled has-property to token me. Token me is labeled ENTITY and has an edge labeled SCENE to token despicable.
Dependency graph for sentence: the tired dog. Token tired is labeled INTERNAL-STATE and has an edge labeled has-state to token dog. Token dog is labeled ENTITY and has an edge labeled SCENE to token tired.
Dependency graph for sentence: Kim ran fast. Token ran is labeled LOCATION-DYN and has an edge labeled has-location to token Kim and an edge labeled SCENE to token fast. Token fast is labeled PROPERTY and has an edge labeled has-property to token ran.
Dependency graph for sentence: Kim ran far. Token ran is labeled LOCATION-DYN and has an edge labeled has-location to token Kim and an edge labeled SCENE to token far. Token far is labeled QUANTITY and has an edge labeled has-quantity to token ran.
Dependency graph for sentence: Ich spiele lieber Schach. Token spiele is labeled META and has an edge labeled participant to token Ich and an edge labeled SCENE to token lieber and an edge labeled scene to token Schach. Token lieber is labeled EVALUATION and has an edge labeled evaluee to token spiele. Token Schach is labeled SITUATION-DYN and has an edge labeled situee to token Ich.
Dependency graph for sentence: Der ist sowieso kaputt. Token sowieso is labeled CONDITION and has an edge labeled has-condition to token kaputt. Token kaputt is labeled INTERNAL-STATE and has an edge labeled has-state to token Der and an edge labeled SCENE to token sowieso.
Dependency graph for sentence: Und doch sah er intelligent aus. Token doch is labeled CONCESSION and has an edge labeled asserted to token sah. Token sah is labeled MESSAGE and has an edge labeled SCENE to token doch and an edge labeled topic to token er and an edge labeled message to token intelligent.
Dependency graph for sentence: Sie sprangen des Regens ungeachtet nach draußen. Token sprangen is labeled LOCATION-INIT and has an edge labeled has-location to token Sie and an edge labeled SCENE to token ungeachtet and an edge labeled LOCATION to token draußen. Token ungeachtet is labeled CONCESSION and has an edge labeled conceded to token Regens.
Dependency graph for sentence: Kim war unvermindert fröhlich. Token unvermindert is labeled QUANTITY-CONTINUATION and has an edge labeled has-quantity to token fröhlich. Token fröhlich is labeled INTERNAL-STATE and has an edge labeled has-state to token Kim and an edge labeled SCENE to token unvermindert.
Dependency graph for sentence: » Wir haben um Hilfe gebeten « , so Saqib. Token gebeten is labeled NORM and has an edge labeled SENDER to token Wir and an edge labeled norm to token Hilfe and an edge labeled SCENE to token so. Token so is labeled SENDING and has an edge labeled sent to token gebeten and an edge labeled sender to token Saqib.
Dependency graph for sentence: ein anderes Plakat. Token anderes is labeled COMPARISON and has an edge labeled compared to token Plakat. Token Plakat is labeled ENTITY and has an edge labeled SCENE to token anderes.
Relative clauses work very similarly: they modify a noun but also act as
predicates assigning that noun a role. Again, we use SCENE for the
acl:relcl relation, and a backlink to the modified noun. This is in addition
to the dependency to any relative pronoun, which has the same role.
Dependency graph for sentence: the song that I like. Token song is labeled ENTITY and has an edge labeled EXPERIENCER to token I and an edge labeled SCENE to token like. Token like is labeled EVALUATION and has an edge labeled evaluee to token song and an edge labeled evaluee to token that.
Dependency graph for sentence: the song I like. Token song is labeled ENTITY and has an edge labeled EXPERIENCER to token I and an edge labeled SCENE to token like. Token like is labeled EVALUATION and has an edge labeled evaluee to token song.