To predict cybersecurity incidents in an IoT environment, transform passive defense into “active” defense, and minimize potential damage to network systems, cybersecurity situational awareness technology has been rapidly developing. This paper integrates common methods for multi-source data fusion, applies an adaptive neural fuzzy inference system to assess cybersecurity situational awareness, and proposes an improved LSTM cybersecurity situational awareness prediction model based on the Sparrow Search Algorithm, achieving cybersecurity protection under multimedia fusion technology. Experimental results show that by quantitatively calculating cybersecurity posture values at different levels—service level, host level, and network system level— this assessment model provides more comprehensive assessment information compared to traditional methods, with more accurate results. Additionally, the average relative error of the improved SSA-LSTM neural network cybersecurity posture prediction algorithm stabilizes around 5.29%, enabling effective prediction of cybersecurity posture over the coming period.
