DASI-Cheff Bot

Project for DASI subject (UCM, course 2019-2020) consisting on a multi-agent system.

The system is developed with SPADE platform, and makes use of a XMPP server.

The main purpose is to develop a multi-agent system able to recognize food and ingredients from pictures and sugguest receipes that could be cooked with such ingredients.

Instalation and dependencies

This software uses Python >= 3.6. You can install packages dependencies with pip as following:

pip install -r requirements.txt

Other considerations:

• There is no need to install any XMPP server if agents names are not modified.
• You will need Dialogflow credentials.
• The neural network model can be downloaded from https://drive.google.com/open?id=19PA-QcdE7IBYzLSc-QPdXHSShHL3_G9y. You must place it in src/cnn_model/saved_model folder.

App execution

1. Add your config.py file with the corresponding credentials to src folder.
2. Execute the script src/main.py.

User's guide

Connect with Telegram Bot

Once the application is running, the user can interact with it via Telegram. There are 2 available options:

1. Accessing link https://t.me/DASIChef_bot
2. Search in the Telegram app for DASIChef_bot.

Bot interaction

Type /start for starting a conversation with the bot.:raw-html-m2r:<br> To finish the conversation, type /exit on the main menu.

Authors and credits

This project is managed by Raul Murillo (ramuri01@ucm.es) and Ignacio Regueiro (iregueir@ucm.es).

Source: https://github.com/RaulMurillo/DASI-SPADE.