University Administration Student Aid Chatbot using RASA

Set-up to build a chatbot in rasa

• Rasa need specific python version, hence create a enviorment using the following command conda create -n rasa-env python=3.8
• now actiavte the enviorment conda activate rasa-env
• now install rasa pip install rasa
• other requirements have been listed in requirements.txt. run pip install -r requirements.txt
• a new project can be initiallized using the command rasa init, rasa train to train the rasa bot, rasa train nlu to only train nlu, rasa run actions to start the actions server with is esential for database contectivity, rasa run to run the bot and rasa shell to open the interactive chatbot. rasa test to test the bost in different case senerios.

NOTE: we have already added the trained model in the repo so to run the bot only enviornment set up is needed then user can do rasa run actions followed by rasa shell

commands to run the bot in UI

• rasa run actions
• rasa run --enable-api --cors "*": This command runs the Rasa server with API enabled and CORS (Cross-Origin Resource Sharing) allowed from all origins, which is necessary for Streamlit to communicate with Rasa.
• streamlit run app.py: Run Streamlit app (Open the URL displayed in terminal to view the Streamlit application)

NOTE: we tried deploying the streamlit bot usingb streamlit sharing, however it was always giving some kind of error. you can check the deployement here https://university-chatbot.streamlit.app/ but it doesnt work. So we advise that the UI be run using the 3 commmand mentioned above.