README

HITT: Hypernymy Identification in Chinese Traffic Legal Text

Download Stanford CoreNLP toolkit and the package for Chinese processing;
Download pre-trained word embedding
Use Stanford CoreNLP toolkit to parse the row sentence data and get the sequence of all the sentences by syntactic parsing;
Config the config.py file, use daojiao_pairs.xlsx file run the context_finder.py file to find the context of each pair in all sentence sequences;
Save the entity pairs, label information together with context information as input data;
Use Stanford CoreNLP to parse the input data into input sequence;
Run following command in all model files:

python train.py <path_to_training_file> <path_to_word_embedding_file>

Name		Name	Last commit message	Last commit date
Latest commit History 22 Commits
CNN		CNN
Pre-processing		Pre-processing
Pre_Training_Bert		Pre_Training_Bert
RNN		RNN
Seq2Seq_Attention		Seq2Seq_Attention
Transformer		Transformer
data		data
fastText		fastText
README.md		README.md