No abstract

System logs record system states and significant events at various critical points to help debug performance issues and failures. Therefore, the rapid and accurate detection of the system log is crucial to the security and stability of the system. In this paper, proposed is a novel attention-based neural network model, which would learn log patterns from normal execution. Concretely, our model adopts a GRU module with attention mechanism to extract the comprehensive and intricate correlations and patterns embedded in a sequence of log entries. Experimental results demonstrate that our proposed approach is effective and achieve better performance than conventional methods.

No abstract

Abstract. In link prediction, the single index prediction effect depends on whether the method can reasonably describe the target network topology characteristics. What's more, the network features are only described from a certain perspective, this limitation leads to the low robustness of a single index for the prediction of different networks. Based on this, we proposed a hybrid method based on Logistic model, considering the similarity index of the prediction results of complementary characteristics and importance in different networks, adaptive fusion gives each index weight reasonably. Experiments show that the AUC and Precision of the hybrid method on each target network are higher than those of the baseline.