This paper presents a collaborative par- titioning algorithm—a novel ensemble- based approach to coreference resolution. Starting from the all-singleton partition, we search for a solution close to the en- semble’s outputs in terms of a task-specific similarity measure. Our approach assumes a loose integration of individual compo- nents of the ensemble and can therefore combine arbitrary coreference resolvers, regardless of their models. Our experi- ments on the CoNLL dataset show that collaborative partitioning yields results su- perior to those attained by the individual components, for ensembles of both strong and weak systems. Moreover, by applying the collaborative partitioning algorithm on top of three state-of-the-art resolvers, we obtain the second-best coreference per- formance reported so far in the literature (MELA v08 score of 64.47).
Collaborative Partitioning for Coreference Resolution / Uryupina, Olga; Moschitti, Alessandro. - ELETTRONICO. - (2017), pp. 47-57. (Intervento presentato al convegno 21st Conference on Computational Natural Language Learning (CoNLL 2017) tenutosi a Vancouver, Canada nel 3-4 August, 2017) [10.18653/v1/K17-1007].
Collaborative Partitioning for Coreference Resolution
Olga Uryupina;Alessandro Moschitti
2017-01-01
Abstract
This paper presents a collaborative par- titioning algorithm—a novel ensemble- based approach to coreference resolution. Starting from the all-singleton partition, we search for a solution close to the en- semble’s outputs in terms of a task-specific similarity measure. Our approach assumes a loose integration of individual compo- nents of the ensemble and can therefore combine arbitrary coreference resolvers, regardless of their models. Our experi- ments on the CoNLL dataset show that collaborative partitioning yields results su- perior to those attained by the individual components, for ensembles of both strong and weak systems. Moreover, by applying the collaborative partitioning algorithm on top of three state-of-the-art resolvers, we obtain the second-best coreference per- formance reported so far in the literature (MELA v08 score of 64.47).File | Dimensione | Formato | |
---|---|---|---|
2017_CoNLL_Moschitti.pdf
accesso aperto
Tipologia:
Post-print referato (Refereed author’s manuscript)
Licenza:
Tutti i diritti riservati (All rights reserved)
Dimensione
221.57 kB
Formato
Adobe PDF
|
221.57 kB | Adobe PDF | Visualizza/Apri |
K17-1007.pdf
accesso aperto
Tipologia:
Versione editoriale (Publisher’s layout)
Licenza:
Creative commons
Dimensione
176.29 kB
Formato
Adobe PDF
|
176.29 kB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione