yuxuan2015 / text-gcn

text_gcn

The implementation of Text GCN in paper:

Liang Yao, Chengsheng Mao, Yuan Luo. "Graph Convolutional Networks for Text Classification." In 33rd AAAI Conference on Artificial Intelligence (AAAI-19)

Dataset

百度云盘链接：https://pan.baidu.com/s/1jraAYlL0CzM6wQThvTCOyQ 提取码：qbtc

Require

Python 2.7 or 3.6

Tensorflow >= 1.4.0

Reproducing Results

Run sh run.sh 20ng 或者 sh run.sh 20ng R8 R52 ohsumed mr

Change 20ng in above 3 command lines to R8, R52, ohsumed and mr when producing results for other datasets.

Example input data

1. /data/20ng.txt indicates document names, training/test split, document labels. Each line is for a document.

2. /data/corpus/20ng.txt contains raw text of each document, each line is for the corresponding line in /data/20ng.txt

Inductive version

An inductive version of Text GCN is fast_text_gcn, where test documents are not included in training process.