Risk Assessment of Security Assets for Networked Swarm Systems Based on Time Series Decomposition of Temporal Knowledge Graph

Assessing the risk of security assets presents a formidable challenge within the realm of cybersecurity. The time series decomposition risk assessment (TSRA), which is a model using time series decomposition techniques, is used in our paper to capture the evolution process of knowledge graph effectively. The evolution process of entities and relations are fitted into multidimensional additive time series, comprising trend components, seasonal components, and residual components. By calculating different weights, long short-term memory(LSTM) is utilized for global information aggregation from historical snapshots. Time series additive decomposition and information aggregation for secure asset risk assessment jointly simulate the evolution of temporal knowledge graphs(TKGs), which allows for dynamic risk assessment of assets in cyberspace.