learning-to-reweight-in-relation-extraction

This is a project that applies the learning strategy in the paper https://arxiv.org/pdf/1803.09050.pdf.

We want to use it in the relation extraction area, hoping to deal with the noisy dataset problem.

Future plan

experiments:

1. noisy training data + PCNN

2. clean training data + PCNN

3. noisy training data + l2rw

Software required

Python 2.7 PyTorch 0.4.0. tensorflow 1.12.0

Dataset

I use the GIDS dataset and modified the PCCN model, thanks to SharmisthaJat: https://github.com/SharmisthaJat/RE-DS-Word-Attention-Models

Running command

python2.7 <file> <data directory> <train file name> <test file name> <dev file name> <word embedding file name>

Acknowledgement

danieltan07: https://github.com/danieltan07/learning-to-reweight-examples#acknowledgements

SharmisthaJat: https://github.com/SharmisthaJat/RE-DS-Word-Attention-Models

yunjey: https://github.com/yunjey/pytorch-tutorial/tree/master/tutorials/04-utils/tensorboard

Citation

Paper Link: https://arxiv.org/abs/1804.06987