Cyber-attack activities are complex and ever-changing, posing severe challenges to cybersecurity personnel. Introducing knowledge graphs into the field of cybersecurity helps depict the intricate cybersecurity landscape and provides technical support for threat identification and situational awareness. However, during the process of constructing knowledge graphs, inevitable noise and conflicts may be introduced, leading to misleading inferences and decisions. This paper aims to address the issues of constructing and assessing the quality of cybersecurity knowledge graphs. We manually constructed a dataset of cybersecurity knowledge graphs (CS13K) and expanded the existing cybersecurity ontology, building a more versatile ontology applicable to existing data features. Building upon this, we utilized Neo4j to construct the cybersecurity knowledge graph. Furthermore, we propose an AttTucker model, based on Transformer, for assessing the quality of knowledge graphs. By employing numerous self-attention heads, this model captures latent information among entities and relations. While reducing the dimensionality of knowledge embeddings, the model achieves evaluation results comparable to high-dimensional embeddings. Finally, we incorporate the path-level information between entities in the knowledge graph into the evaluation of knowledge graph quality based on the AttTucker model. Experimental analysis on both generic datasets and cybersecurity datasets demonstrates that our model significantly improves F1 value and accuracy in knowledge graph quality assessment tasks, outperforming other models.

In brain–computer interface (BCI)-based motor imagery, the symmetric positive definite (SPD) covariance matrices of electroencephalogram (EEG) signals with discriminative information features lie on a Riemannian manifold, which is currently attracting increasing attention. Under a Riemannian manifold perspective, we propose a non-linear dimensionality reduction algorithm based on neural networks to construct a more discriminative low-dimensional SPD manifold. To this end, we design a novel non-linear shrinkage layer to modify the extreme eigenvalues of the SPD matrix properly, then combine the traditional bilinear mapping to non-linearly reduce the dimensionality of SPD matrices from manifold to manifold. Further, we build the SPD manifold network on a Siamese architecture which can learn the similarity metric from the data. Subsequently, the effective signal classification method named minimum distance to Riemannian mean (MDRM) can be implemented directly on the low-dimensional manifold. Finally, a regularization layer is proposed to perform subject-to-subject transfer by exploiting the geometric relationships of multi-subject. Numerical experiments for synthetic data and EEG signal datasets indicate the effectiveness of the proposed manifold network.

Relation classification is a significant task within the field of natural language processing. Its objective is to extract and identify relations between two entities in a given text. Within the scope of this paper, we construct an artificial dataset (CS13K) for relation classification in the realm of cybersecurity and propose two models for processing such tasks. For any sentence containing two target entities, we first locate the entities and fine-tune the pre-trained BERT model. Next, we utilize graph attention networks to iteratively update word nodes and relation nodes. A new relation classification model is constructed by concatenating the updated vectors of word nodes and relation nodes. Our proposed model achieved exceptional performance on the SemEval-2010 task 8 dataset, surpassing previous approaches with a remarkable F1 value of 92.3%. Additionally, we propose the integration of a ranking-based voting mechanism into the existing model. Our best results are an F1 value of 92.5% on the SemEval-2010 task 8 dataset and a value 94.6% on the CS13K dataset. These findings highlight the effectiveness of our proposed models in tackling relation classification tasks.