Ensemble detection model for profile injection attacks in collaborative recommender systems based on BP neural network

摘要: The existing supervised approaches suffer from low precision when detecting profile injection attacks. To solve this problem, the authors propose an ensemble detection model by introducing back propogation (BP) neural network and learning technique. Firstly, through combination of various attack types, they create base training sets which include samples profiles have great diversities with each other. Secondly, use created to train BP networks generate diverse classifiers. Finally, select parts classifiers highest on validation dataset integrate them using voting strategy. Uncorrelated misclassifications generated classifier can be successfully corrected learning. experimental results two different scale real datasets MovieLens Netflix show that proposed effectively improve under condition holding a high recall.