Constituency parsing aims to extract a constituency-based parse tree from a sentence that represents its syntactic structure according to a phrase structure grammar.
Example:
Sentence (S)
|
+-------------+------------+
| |
Noun (N) Verb Phrase (VP)
| |
John +-------+--------+
| |
Verb (V) Noun (N)
| |
sees Bill
Recent approaches convert the parse tree into a sequence following a depth-first traversal in order to be able to apply sequence-to-sequence models to it. The linearized version of the above parse tree looks as follows: (S (N) (VP V N)).
The Wall Street Journal section of the Penn Treebank is used for evaluating constituency parsers. Section 22 is used for development and Section 23 is used for evaluation. Models are evaluated based on F1. Most of the below models incorporate external data or features. For a comparison of single models trained only on WSJ, refer to Kitaev and Klein (2018).
| Model | F1 score | Paper / Source | Code |
|---|---|---|---|
| Span Attention + XLNet (Tian et al., 2020) | 96.40 | Improving Constituency Parsing with Span Attention | Official |
| Label Attention Layer + HPSG + XLNet (Mrini et al., 2020) | 96.38 | Rethinking Self-Attention: Towards Interpretability for Neural Parsing | Official |
| Attach-Juxtapose Parser + XLNet (Yang and Deng, 2020) | 96.34 | Strongly Incremental Constituency Parsing with Graph Neural Networks | Official |
| Head-Driven Phrase Structure Grammar Parsing (Joint) + XLNet (Zhou and Zhao, 2019) | 96.33 | Head-Driven Phrase Structure Grammar Parsing on Penn Treebank | |
| Head-Driven Phrase Structure Grammar Parsing (Joint) + BERT (Zhou and Zhao, 2019) | 95.84 | Head-Driven Phrase Structure Grammar Parsing on Penn Treebank | |
| CRF Parser + BERT (Zhang et al., 2020) | 95.69 | Fast and Accurate Neural CRF Constituency Parsing | Official |
| Self-attentive encoder + ELMo (Kitaev and Klein, 2018) | 95.13 | Constituency Parsing with a Self-Attentive Encoder | Official |
| Model combination (Fried et al., 2017) | 94.66 | Improving Neural Parsing by Disentangling Model Combination and Reranking Effects | |
| LSTM Encoder-Decoder + LSTM-LM (Takase et al., 2018) | 94.47 | Direct Output Connection for a High-Rank Language Model | |
| LSTM Encoder-Decoder + LSTM-LM (Suzuki et al., 2018) | 94.32 | An Empirical Study of Building a Strong Baseline for Constituency Parsing | |
| In-order (Liu and Zhang, 2017) | 94.2 | In-Order Transition-based Constituent Parsing | |
| CRF Parser (Zhang et al., 2020) | 94.12 | Fast and Accurate Neural CRF Constituency Parsing | Official |
| Semi-supervised LSTM-LM (Choe and Charniak, 2016) | 93.8 | Parsing as Language Modeling | |
| Stack-only RNNG (Kuncoro et al., 2017) | 93.6 | What Do Recurrent Neural Network Grammars Learn About Syntax? | |
| RNN Grammar (Dyer et al., 2016) | 93.3 | Recurrent Neural Network Grammars | |
| Transformer (Vaswani et al., 2017) | 92.7 | Attention Is All You Need | |
| Combining Constituent Parsers (Fossum and Knight, 2009) | 92.4 | Combining constituent parsers via parse selection or parse hybridization | |
| Semi-supervised LSTM (Vinyals et al., 2015) | 92.1 | Grammar as a Foreign Language | |
| Self-trained parser (McClosky et al., 2006) | 92.1 | Effective Self-Training for Parsing |