Skip to content

Open-source code for our NAACL-2021 paper: "Neural Sequence Segmentation as Determining the Leftmost Segments".

Notifications You must be signed in to change notification settings

LeePleased/LeftmostSeg

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

6 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Leftmost Segment Determination for Chunking

In NAACL-2021, we introduce a fast and effective approach for sequence segmentation tasks (e.g., Chinese POS tagging). This repo. contains the main implementation of our paper.

Setup

Two steps. Firstly, create a folder named "dataset" (containing {train, dev, & test}.txt) and the data format is

中      NR
美      NR
在      P
沪      NR
签订    VV
高      JJ
科技    NN
合作    NN
协议    NN

Secondly, download evaluation script and rename it as "conlleval.pl".

Training and Test

CUDA_VISIBLE_DEVICES=0 python main.py \
    --data_dir dataset \
    --check_dir save \
    --script_path conlleval.pl

Citation

@inproceedings{li-etal-2021-neural,
    title = "Neural Sequence Segmentation as Determining the Leftmost Segments",
    author = "Li, Yangming and Liu, Lemao and Yao, Kaisheng",
    booktitle = "Proceedings of the 2021 Conference of the North {A}merican Chapter of the Association for Computational Linguistics",
    publisher = "Association for Computational Linguistics",
}

About

Open-source code for our NAACL-2021 paper: "Neural Sequence Segmentation as Determining the Leftmost Segments".

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages