| M95-1015 | constructions are a source of error in | appositive recognition | . In addition , the parser confuses |
| X93-1022 | tagging , and coreference based on | appositive recognition | . But we encountered substantial |
| M93-1023 | tagging , and coreference based on | appositive recognition | . But we encountered substantial |
| X93-1022 | the coreference interface for | appositive recognition | , we eliminated from the training |
| M93-1023 | the coreference interface for | appositive recognition | , we eliminated from the training |
| W14-3407 | set/instance anaphora resolution and | appositive recognition | can play a significant role in |
| X93-1022 | trained a classifier to handle | appositive recognition | using EJV development texts and |
| M93-1023 | trained a classifier to handle | appositive recognition | using EJV development text s |
| X93-1022 | noun phrase termination , ( 2 ) | appositive recognition | , and ( 3 ) coreference resolution |
| H92-1063 | prepositional phrase attachment 2 , | appositive recognition | , and on correcting some errors |
| X93-1022 | conjunctions and prepositional phrases . | Appositive recognition | is a subset of the general coreference |
| M91-1006 | prepositional phrase attachment3 , | appositive recognition | , and correcting errors made |
| X93-1022 | led to a mistake in trainable | appositive recognition | , which then combined with a |
| M93-1023 | led to a mistake in trainable | appositive recognition | , which then combine d with a |
| X93-1022 | unrelated ( but adjacent ) NPs . | Appositive recognition | is therefore treated as a binary |
| M93-1023 | unrelated ( but adjacent ) NPs . | Appositive recognition | is therefore treated as a binary |
