All Projects → DeepLearnXMU → RRWEL

DeepLearnXMU / RRWEL

Licence: other
Neural Collective Entity Linking Based on Recurrent Random Walk Network Learning. Code from IJCAI 2019 paper.

Programming Languages

python
139335 projects - #7 most used programming language
HTML
75241 projects
assembly
5116 projects
c
50402 projects - #5 most used programming language
Roff
2310 projects
Jasmin
10 projects
Nemerle
15 projects

Neural Collective Entity Linking Based on Recurrent Random Walk Network Learning

Models and results can be found at our IJCAI 2019 paper [Neural Collective Entity Linking Based on Recurrent Random Walk Network Learning]. It achieves the state-of-the-art result on EL task.

Details will be updated soon.

Requirement:

Python: 3.6.3
PyTorch: 0.3.1 

Input format:

We transform the original data into pkl format, if you want the tranform code, please concat me.

pkl data location:

link:https://pan.baidu.com/s/17tHxyLAMqdOTozmnQsMQ6w

Fetch Code:kwvn

How to run the code?

Local:

python pre_net_xmg.py --cuda_device 0 --nohup regular --epoch 25 --weight_decay 1.28e-5 --LR 0.001 --batch 500 --filter_num 64 --filter_window 3  --local_model_loc model_loc/local/local_regular_new1 --embedding_finetune 1

Global:

python net_global_train.py --cuda_device 0 --nohup 0.5_0.1_3 --weight_decay 1.28e-5 --LR 0.0005 --local_model_loc model_loc/local/local_regular_new1.938.pkl --global_model_loc model_loc/global/global_model --random_k 3 --lamda 0.5 --flag 4:3:1 --gama 0.1 --batch 200 --epoch 25

Cite:

Please cite our IJCAI 2019 paper:

@article{xue2018,  
 title={Neural Collective Entity Linking Based on Recurrent Random Walk Network Learning },  
 author={Mengge Xue, Weiming Cai, Jinsong Su and Linfeng Song, Yubin Ge, Yubao Liu, Bin Wang},  
 booktitle={The Program Committee of the 28th International Joint Conference on Artificial Intelligence (IJCAI-19)},
 year={2019}  
}
Note that the project description data, including the texts, logos, images, and/or trademarks, for each open source project belongs to its rightful owner. If you wish to add or remove any projects, please contact us at [email protected].