Information-Extraction-Chinese

Why we included this project

For anyone doing information extraction on Chinese text, this repo gives you two working deep learning pipelines: a named entity recognizer that combines IDCNN or biLSTM with a CRF layer, and a relation extractor built on biGRU with character and sentence level attention. You get runnable code, not just architecture diagrams, so it is a practical base for teams that need to pull entities like people, organizations, and locations out of documents and link them into typed relations. Each subtask has its own README that walks through data preparation, training, and inference, which helps when you want to adapt the models to your own domain data. It also works well as a teaching reference for how classic sequence labeling and attention behave on Chinese input, especially if you are not using large pretrained language models.