This is the Chatbot_22

The 3 clones are:

1. cafe
2. gym
3. library

The 3 Agents are:

1. Agent 1: GUI and Conversation Context
2. Agent 2: Dialog_act classification and Intent classification
3. Agent 3: Google Search and LDA

Codes were written by:

1. Agent 1: GUI and Connection of Python to the web is written by Mohammad Sina Kiarostami
2. Agent 1: Conversation context and text preprocessing by Hamza Abdalla
3. Agent 2 (Training, deployment), integration of all 3 agents and deployment into web by Aayush Kafle
4. Agent 3: Google Search and LDA by Micheal Klassen

Additional responsibilities:

- Group Leader: Aayush Kafle
- Documentation and Report Preperation: Mohammad Sina Kiarostami
- Midterm Presentation: Micheal Klassen
- Testing, Multiple Clones and Proof reading: Hamza Abdalla

The codes for different agents can be found at:

• Agent 1, GUI and Multiple clones:
  • server_<clone_name>.py, Agent_One_<clone_name>.py, and public_<clone_name>
• Agent 2:
  • Training codes at others/, models at models/, and working codes inside chatbot/
• Agent 3:
  • Agent 3 codes are at chatbot/
• The integration code is written within the agents.

Steps to run (For local computer)

• First create a python3.8 virtual environment and activate it
• Then, in the project root directory pip install -r requirements.txt
• To train, run train.sh script
• Then start rasa server by rasa_run.sh
• You will also need other classification model data. Please contact aayush.kafle@gmail.com for the data.
• Then in another terminal start the respective chatbot clone by chatbot_<clone_name>.sh
• Then go to the address: :/chatbot.html and run respective clone.