Advancing recommender systems by mitigating shilling attacks

This repository contains the code of our take on solving the problem of Shilling attacks.

We achieved the following tasks in our project:

1. Designed multiple recommendation models based on User-User CF,Item-Item CF and Content based filtering.
2. Performed various shilling attacks on each of the model to manipulate their behaviour and analyse the bias introduced.
3. Modelled a detection algorithm to identify and remove the shilling attacker profiles.

I presented our research at the 9th ICCCNT conference held in IISc Bengaluru. Link to our research paper.