About this repo

This repo hosts code for "Transition-based Bubble Parsing: Improvements on Coordination Structure Prediction" (Shi and Lee, ACL 2021).

Running the code

See scripts under scripts/ for example training scripts, and see test.py for example inferencing script.

Some available parameters:

Parameter	Description
wdims	Word embedding dimension
cdims	Character embedding dimesion
edims	External word embedding (e.g., GloVe) dimension
pdims	POS tag embedding dimension
idims	Indicator embedding dimension
bilstm-dims	Bi-LSTM hidden dimension
bilstm-layers	Bi-LSTM layers
bilstm-dropout	Bi-LSTM dropout
char-hidden	Char LSTM dimension (always 1 layer)
char-dropout	Char LSTM dropout
parser-dims	Parser MLP hidden dimension
parser-dropout	Parser MLP dropout
stack-fts	Number of positional features taken from stack
rescore	True or False, enabling/disabling the rescoring module
bert	True or False, using pre-trained BERT features or not

Reference

@InProceedings{shi-lee-21-transition,
    title = "Transition-based Bubble Parsing: Improvements on Coordination Structure Prediction"
    author = "Shi, Tianze  and
              Lee, Lillian",
    booktitle = "Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics",
    month = aug,
    year = "2021",
    address = "Online",
    pages = "7167--7182",
    publisher = "Association for Computational Linguistics",
}