Sentence structural and semantic information are important for relation extraction. This program provides a method of obtaing sentence information based on feature engineering and neural networks. Main content is utilizing entity-related features to construct combined feature.The neural network model we used in our research is BERT model.
The steps for running:
1.Excuting "python main_train.py" will start a trainning programming.
2.Excuting "python main_predict.py" will start a predicting programming.
Or excuting "python Excute_main_predict.py" will start an automic programming for predicting under the newest model.