Feature Optimization for Constituent Parsing via Neural Networks
By Zhiguo Wang, Haitao Mi, Nianwen Xue. 2015
20 citations recorded by the source corpus.
The performance of discriminative constituent parsing relies crucially on feature engineering, and effective features usually have to be carefully selected through a painful manual process. In this paper, we propose to automatically learn a set of effective features via neural networks. Specifically, we build a feedforward neural network model, which takes as input a few primitive units (words, POS tags and certain contextual tokens) from the local context, induces the feature representation in the hidden layer and makes parsing predictions in the output layer. The network simultaneously learns the feature representation and the prediction model parameters using a back propagation algorithm. By pre-training the model on a large amount of automatically parsed data, and then fine-tuning on the manually annotated Treebank data, our parser achieves the highest F1 score at 86.6% on Chinese Treebank 5.1, and a competitive F1 score at 90.7% on English Treebank. More importantly, our parser generalizes well on cross-domain test sets, where we significantly outperform Berkeley parser by 3.4 points on average for Chinese and 2.5 points for English.
Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Natural Language Processing
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Noah A. Smith
- Chris Dyer
- Graham Neubig
- Miguel Ballesteros
- Yoav Goldberg
- A. Kuncoro
- Mohit Bansal
- Yejin Choi
In Computer Vision and Pattern Recognition
- Noah A. Smith
- Chris Dyer
- Miguel Ballesteros
- A. Kuncoro
- Jonathan Berant
- Hannaneh Hajishirzi
- Luke S. Zettlemoyer
- Yonatan Belinkov
In Natural Language Processing
- Chris Dyer
- Noah A. Smith
- Miguel Ballesteros
- Graham Neubig
- Yoav Goldberg
- A. Kuncoro
- Kyunghyun Cho
- Yonatan Belinkov
Citations in the corpus
- Recurrent Neural Network Grammars (2016, methodology): “esults. Chinese parsing results were obtained with the same methodology as in English and show the same pattern (Table 6). Table 3: Parsing results on CTB 5.1. Model type F 1 Zhu et al. (2013) D 82.6 Wang et al. (2015) D 83.2 Huang and Harper (2009) D 84.2 Charniak (2000) G 80.8 Bikel (2004) G 80.6 ”
- In-Order Transition-based Constituent Parsing (2017, unknown): “The in-order parser achieves the best results under hte fully-Model 86.9 Top-down parser 86.9 Bottom-up parser 87.5 In-order parser 88.0 semi-supervision Zhu et al. (2013) 85.6 Wang and Xue (2014) 86.3 Wang et al. (2015) 86.6 Table 6: Final results on test set of CTB. supervised setting.”