Rasa_NLU_Chi

Why we included this project

For teams building Chinese-language assistants, this fork of Rasa NLU provides working pipelines that combine MITIE and Jieba tokenization, with an optional scikit-learn intent classifier. The project includes a pretrained word-feature extractor trained on Chinese Wikipedia and Baidu Baike, plus demo training data, so you can test intent and entity extraction right away. When your chatbot uses specialized terminology, you can add Jieba user dictionaries so those terms are not split incorrectly. If you already know the classic Rasa NLU workflow, this is a direct way to get the understanding layer working without assembling the components yourself.