diff --git a/README.md b/README.md
index 6634bec..60ce29a 100644
--- a/README.md
+++ b/README.md
@@ -1,26 +1,82 @@
+# rasa_chatbot
+A Chinese task oriented chatbot in  IVR(Interactive Voice Response) domain(电信ivr领域， 中文)， Implement by rasa nlu and rasa core. This is a demo with toy dataset.
 
-## install dependency
+### install dependency:
+- [follow here](https://github.com/zqhZY/rasa_chatbot/blob/master/INSTALL.md)
 
-#### python3
-install or update to python 3
+### dir tree
+```
+rasa_chatbot/
+├── data
+│   ├── mobile_nlu_data.json  # rasa nlu train data
+│   ├── mobile_story.md       # rasa core train data
+│   └── total_word_feature_extractor.dat  # mitie word vector feature
+├── __init__.py               # init file
+├── httpserver.py             # rasa nlu httpserver
+├── bot.py                    # ivr bot main script.
+├── mobile_domain.yml         # rasa core domain file
+├── mobile_nlu_model_config.json  # rasa nlu config file
+├── models                    # directory to save trained models
+└── README.md                 # readme file
+
+```
+
+### train nlu model
+```bash
+python bot.py train-nlu
+```
 
-#### install chinese version of rasa nlu
+### train dialogue
+```bash
+python bot.py train-dialogue
 ```
-git clone https://github.com/crownpku/Rasa_NLU_Chi.git
-cd rasa_nlu
-pip install -r requirements.txt
-python setup.py install
+### train dialogue in online mode
+```
+python bot.py online_train
 ```
 
-#### install sklearn and MITIE
+### test
+Run command below:
+```bash
+python bot.py run
+```
+Example1:
+```
+Bot loaded. Type a message and press enter : 
+YOU：你是谁
+BOT：您好!，我是机器人小热，很高兴为您服务。
+YOU：我想看一下消费情况
+BOT：您想查询哪个月的消费？
+YOU：十月份的
+BOT：好，请稍等
+BOT：您好，您十月份共消费二十八元。
+YOU：好谢谢
+BOT：您开心我就开心
+BOT：您还想干什么
+YOU：没什么了
+BOT：Bye， 下次再见
 
 ```
-pip install -U scikit-learn sklearn-crfsuite
-pip install git+https://github.com/mit-nlp/MITIE.git
+Example2:
+```
+Bot loaded. Type a message and press enter : 
+YOU：给我看看我上月用了多少话费
+BOT：好，请稍等
+BOT：您好，您上月共消费二十八元。
+BOT：您还想干什么
+
 ```
 
-#### install rasa_core
+### train word vector
 
+You can train your own MITIE model using following method:
 ```
-pip install rasa_core
+$ git clone https://github.com/mit-nlp/MITIE.git
+$ cd MITIE/tools/wordrep
+$ mkdir build
+$ cd build
+$ cmake ..
+$ cmake --build . --config Release
+$ ./wordrep -e /path/to/your/folder_of_cutted_text_files
 ```
+/path/to/your/folder_of_cutted_text_files above is a directory path in which has word cutted data files to train. This process may cost one or two days.