A METHOD FOR DETECTING MANIPULATION ATTACKS ON RECOMMENDATION SYSTEMS WITH COLLABORATIVE FILTERING

Information security application
Authors:
ZAGALSKII D. S. Solovey R. S. Dakhnovich A. D.
Abstract:

This article discusses the security of recommendation systems with collaborative filtering
from manipulation attacks. The most common types of attacks are analyzed and identified. A modified
method for detecting manipulation attacks on recommendation systems with collaborative
filtering is proposed. Experimental testing and comparison of the effectiveness of the modified
method with other relevant methods were carried out