Parsing as pretraining
Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review
Documents
- Parsing as pretraining
Accepted author manuscript, 819 KB, PDF document
Recent analyses suggest that encoders pretrained for language modeling capture certain morpho-syntactic structure. However, probing frameworks for word vectors still do not report results on standard setups such as constituent and dependency parsing. This paper addresses this problem and does full parsing (on English) relying only on pretraining architectures – and no decoding. We first cast constituent and dependency parsing as sequence tagging. We then use a single feed-forward layer to directly map word vectors to labels that encode a linearized tree. This is used to: (i) see how far we can reach on syntax modelling with just pretrained encoders, and (ii) shed some light about the syntax-sensitivity of different word vectors (by freezing the weights of the pretraining network during training). For evaluation, we use bracketing F1-score and las, and analyze in-depth differences across representations for span lengths and dependency displacements. The overall results surpass existing sequence tagging parsers on the ptb (93.5%) and end-to-end en-ewt ud (78.8%).
Original language | English |
---|---|
Title of host publication | Proceedings of the AAAI Conference on Artificial Intelligence (AAAI 2020) |
Publisher | AAAI Press |
Publication date | 2020 |
Pages | 9114-9121. |
ISBN (Electronic) | 978-1-57735-835-0 |
DOIs | |
Publication status | Published - 2020 |
Event | Thirty-Forth AAAI Conference on Artificial Intelligence: AAAI 2020 - New York, United States Duration: 7 Feb 2020 → 12 Feb 2020 https://aaai.org/Conferences/AAAI-20/ |
Conference
Conference | Thirty-Forth AAAI Conference on Artificial Intelligence |
---|---|
Land | United States |
By | New York |
Periode | 07/02/2020 → 12/02/2020 |
Internetadresse |
ID: 258333711