Neural PP Attachment Disambiguation Systems

Dataset

DOI

This resource contains code for different types of neural PP attachment disambiguation systems:

A disambiguation system inspired by de Kok et al. (2017) but with the ranking loss function. A disambiguation system with biaffine attention similar to the neural dependency parser in Dozat and Manning (2017).

The systems are described in details in the paper: Do and Rehbein (2020). "Parsers Know Best: German PP Attachment Revisited". We also include all pre-trained models reported in the paper.

Identifier
DOI	https://doi.org/10.11588/data/DKWKGJ
Related Identifier	https://doi.org/10.18653/v1/2020.coling-main.185
Metadata Access	https://heidata.uni-heidelberg.de/oai?verb=GetRecord&metadataPrefix=oai_datacite&identifier=doi:10.11588/data/DKWKGJ

Provenance
Creator	Do, Bich-Ngoc; Rehbein, Ines
Publisher	heiDATA
Contributor	Do, Bich-Ngoc
Publication Year	2023
Rights	info:eu-repo/semantics/openAccess
OpenAccess	true
Contact	Do, Bich-Ngoc (Institute of Computational Linguistics, Heidelberg University & Leibniz Institute for German Language)

Representation
Resource Type	source code; Dataset
Format	application/zip; application/gzip; text/markdown
Size	54286; 62364967; 162960282; 193821425; 194933783; 164898136; 193738237; 164589547; 59588; 9293326; 9461492; 13905299; 15112221; 2260; 285; 1286
Version	1.0
Discipline	Humanities