Empathy Engine with Microsoft LUIS

This is the backend server for the conversational AI used in the implementation of Flood Disaster VR

New Features!

• Handles converstaions as POST requests as a REST service
• Uses LUIS to process the NLU
• Handles NLP sentiment analysis

Tech

This project is built using Python 3.6.1

Conversation Training

The LUIS AI is trained on intents around the following. Feel free to speak around them to get the objective complete. You dont have to speak the exact sentences, just on the same context.

• Hi/Hello
• What is your name?
• How old are you?
• Are you okay?
• I'm here to help
• Where is your family?
• Do you live here?
• What happened?
• Are you hungry?
• Are you injured?
• Can you move?
• Come with me

LUIS Intents

The intents for LUIS are given as a json file that can be imported into LUIS. It is the file called luis_intent_import.json present in the root directory of this repository.

Installation

This repository requires Python v3.6 to run.

Install the dependencies and devDependencies and start the server running on PORT 3000

$ cd empathy-engine-luis

Create a .env file and put the keys and endpoints for LUIS as LUIS_KEY and LUIS_ENDPOINT

$ pip install -r requirements.txt
$ python app.py

Endpoints

• POST - /ml/api/v1.0/assistant Request: { "sentence":"the text to be sent" }

• GET - /ml/api/v1.0/reset

`