Milktea Chinese Chatbot

This project uses the rasa framework to provide a chatbot for ordering boba tea, taking 幸福堂 as a reference.

Live system

You may access a live version of this chatbot via our server at home.

Setting up

System requirements

• System with support for the x86 instruction set
• Git
• Docker
• NodeJS (with npm)

Scripted setup (Linux systems)

1. Clone the repo from github

git clone https://github.com/Skyquek/milktea-CN_chatbot-nlp.git

2. Give exectable permissions

cd milktea-CN_chatbot-nlp
chmod 0777 buildAll.sh

3. Run the build process

./buildAll.sh

4. Once everything has finished building, you may access the chatbot service via your browser at:

http://<your-ip>:52288/

Non-scripted setup

1. Clone the repo from github, then go into the folder

git clone https://github.com/Skyquek/milktea-CN_chatbot-nlp.git
cd milktea-CN_chatbot-nlp

Build the frontend

1. Go into the folder containing the frontend code

cd ./public

2. Build the frontend files

npm install
ng build --prod

3. Build the docker image

docker build -t xingfutang-chatbot-frontend  . --no-cache

4. Run the docker image (you may change the port number - 52288 to any number you desire)

docker run --name xingfutang-chatbot-frontend  -p 52288:80 -d xingfutang-chatbot-frontend

5. Done!

Build the backend

1. Build the docker image

docker build -t nlp-chatbot:latest .

2. Run the docker image

docker run --name xinfutang -p 5005:5005 -d ai-trademark-frontend

3. Done!

Building without docker

1. Start bert-as-service. bert-as-service is a sentence encoding service for mapping a variable-length sentence to a fixed-length vector.

bert-serving-start -model_dir chinese_L-12_H-768_A-12/ -num_worker=1

2. start action server

rasa run actions

3. train a model

rasa train

4. talk to your chatbot

   • API: rasa run

   • CMD: rasa shell

5. EXPOSE the PORT NUMBER

URL: "http://localhost:5005/webhooks/rest/webhook"
Method: POST
Header: Content-Type: application/json
Body:
{
"sender": "Quek",
"message": “奶茶”
}
Response:
[
  {
      "recipient_id": "Quek",
      "text": "你好"
  },
  {
      "recipient_id": "Quek",
      "text": "想要喝点什么？"
  }
]