📕 AI Module built for BiblioNet. 📕
A project for
Fundamentals of Artificial Intelligence
course of Computer Science at University of Salerno.
📕 🎓 Our module's goal is to help new subscribers find a genre to start with. We let the user answer to 5 questions with a quiz, and then our AI will suggest genres that they could like. We achieved this with a machine learning agent, using Python. In particular we used parameter learning with Bayesian Model. The training dataset was built from scratch, making people answer our questions with an online form.
Process documentation can be found in documents directory. The AI module has always been in our plans for Biblionet. You can find documentation about requirement analysis and system design at the following link
- Stefano Lambiase - Project Manager - StefanoLambiase
- Gianmario Voria - Developer - gianwario
- Viviana Pentangelo - Developer - vipenti
- Nicola Pagliara - Developer - Nicola-Pagliara
In this section we introduce technical informations and installing guides!
First of all, you need to install Biblionet. Follow the instruction at this link to install the project, but do it with this repository, as it contains the module.
We assume you have followed previous instructions and that you have a correct installation of Python on your PC. Under the folder /src/main/java/it/unisa/c07/biblionet/moduloIntelligenzaArtificiale you should find a file named "model.obj", which contains our machine learning agent already trained. You can now run the script named "prediction.py". This will create an API running on "localhost:5000" that the application is going to use to make predictions.
Run Biblionet, register as "Lettore", go into your user area and press on "Questionario di Supporto" button.