We present Køpsala, the Copenhagen-Uppsala system for the Enhanced Universal Dependencies Shared Task at IWPT 2020. Our system is a pipeline consisting of off-the-shelf models for everything but enhanced graph parsing, and for the latter, a transition-based graph parser adapted from Che et al. (2019). We train a single enhanced parser model per language, using gold sentence splitting and tokenization for training, and rely only on tokenized surface forms and multilingual BERT for encoding. While a bug introduced just before submission resulted in a severe drop in precision, its post-submission fix would bring us to 4th place in the official ranking, according to average ELAS. Our parser demonstrates that a unified pipeline is effective for both Meaning Representation Parsing and Enhanced Universal Dependencies.
Daniel Hershcovich, Miryam de Lhoneux, Artur Kulmizev, Elham Pejhan, and Joakim Nivre. 2020.Køpsala: Transition-Based Graph Parsing via Efficient Training and Effective Encoding. InProceedings of the 16th International Conference on Parsing Technologies and the IWPT 2020 Shared Task on Parsing into Enhanced Universal Dependencies, pages 236–244, Online. Association for Computational Linguistics.
@inproceedings{hershcovich-etal-2020-kopsala, title = "{K}{\o}psala: Transition-Based Graph Parsing via Efficient Training and Effective Encoding", author = "Hershcovich, Daniel and de Lhoneux, Miryam and Kulmizev, Artur and Pejhan, Elham and Nivre, Joakim", editor = "Bouma, Gosse and Matsumoto, Yuji and Oepen, Stephan and Sagae, Kenji and Seddah, Djam{\'e} and Sun, Weiwei and S{\o}gaard, Anders and Tsarfaty, Reut and Zeman, Dan", booktitle = "Proceedings of the 16th International Conference on Parsing Technologies and the IWPT 2020 Shared Task on Parsing into Enhanced Universal Dependencies", month = jul, year = "2020", address = "Online", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2020.iwpt-1.25/", doi = "10.18653/v1/2020.iwpt-1.25", pages = "236--244", abstract = "We present K{\o}psala, the Copenhagen-Uppsala system for the Enhanced Universal Dependencies Shared Task at IWPT 2020. Our system is a pipeline consisting of off-the-shelf models for everything but enhanced graph parsing, and for the latter, a transition-based graph parser adapted from Che et al. (2019). We train a single enhanced parser model per language, using gold sentence splitting and tokenization for training, and rely only on tokenized surface forms and multilingual BERT for encoding. While a bug introduced just before submission resulted in a severe drop in precision, its post-submission fix would bring us to 4th place in the official ranking, according to average ELAS. Our parser demonstrates that a unified pipeline is effective for both Meaning Representation Parsing and Enhanced Universal Dependencies."}
%0 Conference Proceedings%T Køpsala: Transition-Based Graph Parsing via Efficient Training and Effective Encoding%A Hershcovich, Daniel%A de Lhoneux, Miryam%A Kulmizev, Artur%A Pejhan, Elham%A Nivre, Joakim%Y Bouma, Gosse%Y Matsumoto, Yuji%Y Oepen, Stephan%Y Sagae, Kenji%Y Seddah, Djamé%Y Sun, Weiwei%Y Søgaard, Anders%Y Tsarfaty, Reut%Y Zeman, Dan%S Proceedings of the 16th International Conference on Parsing Technologies and the IWPT 2020 Shared Task on Parsing into Enhanced Universal Dependencies%D 2020%8 July%I Association for Computational Linguistics%C Online%F hershcovich-etal-2020-kopsala%X We present Køpsala, the Copenhagen-Uppsala system for the Enhanced Universal Dependencies Shared Task at IWPT 2020. Our system is a pipeline consisting of off-the-shelf models for everything but enhanced graph parsing, and for the latter, a transition-based graph parser adapted from Che et al. (2019). We train a single enhanced parser model per language, using gold sentence splitting and tokenization for training, and rely only on tokenized surface forms and multilingual BERT for encoding. While a bug introduced just before submission resulted in a severe drop in precision, its post-submission fix would bring us to 4th place in the official ranking, according to average ELAS. Our parser demonstrates that a unified pipeline is effective for both Meaning Representation Parsing and Enhanced Universal Dependencies.%R 10.18653/v1/2020.iwpt-1.25%U https://aclanthology.org/2020.iwpt-1.25/%U https://doi.org/10.18653/v1/2020.iwpt-1.25%P 236-244
[Køpsala: Transition-Based Graph Parsing via Efficient Training and Effective Encoding](https://aclanthology.org/2020.iwpt-1.25/) (Hershcovich et al., IWPT 2020)
Daniel Hershcovich, Miryam de Lhoneux, Artur Kulmizev, Elham Pejhan, and Joakim Nivre. 2020.Køpsala: Transition-Based Graph Parsing via Efficient Training and Effective Encoding. InProceedings of the 16th International Conference on Parsing Technologies and the IWPT 2020 Shared Task on Parsing into Enhanced Universal Dependencies, pages 236–244, Online. Association for Computational Linguistics.