Recommendation attack detection based on deep learning

Virtual Reality and Internet of Things-Based Music Online Learning via the Graph Neural Network

Virtual reality and the Internet of Things have shown their capability in a variety of tasks. However, their availability in online learning remains an unresolved issue. To bridge this gap, we propose a virtual reality and Internet of Things-based pipeline for online music learning. The one graph network is used to generate an automated evaluation of learning performance which traditionally was given by the teachers. To be specific, a graph neural network-based algorithm is employed to identify the real-time status of each student within an online class. In the proposed algorithm, the characteristics of each student collected from the multisensors deployed on their bodies are taken as the input feature for each node in the presented graph neural network. With the adoption of convolutional layers and dense layers as well as the similarity between each pair of students, the proposed approach can predict the future circumstance of the entire class. To evaluate the performance of our work, comparison experiments between several state-of-the-art algorithms and the proposed algorithm were conducted. The result from the experiments demonstrated that the graph neural network-based algorithm achieved competitive performance (sensitivity 91.24%, specificity 93.58%, and accuracy 89.79%) over the state-of-the-art.

Detecting Shilling Attacks Using Hybrid Deep Learning Models

Recommendation systems play a significant role in alleviating information overload in the digital world. They provide suggestions to users based on past symmetric activities or behaviors. Being heavily dependent on users’ behavior, they tend to be vulnerable to shilling attacks. Therefore, protecting them from attacks’ effects is highly important. As shilling attacks have features of a large number of ratings and increasing complexity in attack models, deep learning methods become proper alternatives for more accurate attack detections. This paper proposes a hybrid model of two different neural networks, convolutional and recurrent neural networks, to detect shilling attacks efficiently. The proposed deep learning model utilizes the transformed network architecture for undertaking the attributes derived from user-rated profiles. This architecture enables modeling of the temporal and spatial information in the recommendation system’s ratings. The hybrid model overcomes the limitations of the existing shilling attack deep-learning methods to enhance the recommendation systems’ efficiency and robustness. Experimental results show that the hybrid model results in better predictions on the Movie-Lens 100 K and Netflix datasets by accurately detecting most of the obfuscated attacks compared to the state-of-art deep learning algorithms used for investigation.