MultiChannel

Multi-channel LSTM-CNN model for Vietnamese sentiment analysis

This is implementation for the paper "Multi-channel LSTM-CNN model for Vietnamese sentiment analysis" [link].

We provide a sentiment analysis dataset: VS. The dataset include two version: tokenized and without tokenized.

To run this code:

1. Please specify a data path in preprocessing.py and load_data.py.
2. Run "python preprocessing.py" and then "python cnn_lstm.py"

Requirement:

1. Keras
2. Tensorflow

Contributing

Please create an issue at https://github.com/ntienhuy/MultiChannel reposiroty. We encourage people to contribute to this repository and apply in practical application.

License

• The code in this repository is released under the terms of the GNU General Public License v3.0.
• If you use our data or implementation, please cite our following paper and share your idea by email :

@INPROCEEDINGS{Quan, 
	author={Q. H. Vo and H. T. Nguyen and B. Le and M. L. Nguyen}, 
	booktitle={2017 9th International Conference on Knowledge and Systems Engineering (KSE)}, 
	title={Multi-channel LSTM-CNN model for Vietnamese sentiment analysis}, 
	year={2017},  
	pages={24-29}, 
	keywords={Analytical models;Feature extraction;Logic gates;Machine learning;Neural networks;Sentiment analysis}, 
	doi={10.1109/KSE.2017.8119429}, 
	month={Oct},
}

Contact

Nguyen Tien Huy: ntienhuy@jaist.ac.jp