Managing emergency crises using secure information through educational awareness: COVID-19 case study.

Social networks are increasingly taking over daily life, creating a volume of unsecured data and making it very difficult to capture safe data, especially in times of crisis. This study aims to use a Convolutional Neural Network (CNN)-Long Short-Term Memory (LSTM)-based hybrid model for health monitoring and health crisis forecasting. It consists of efficiently retrieving safe content from multiple social media sources. Educational awareness is a fairly important tool and a constant reminder to do everything to avoid fake news. The hybrid model captures safe and meaningful features from multiple social media sources. This research study enables retrieval of qualitative and secure content and mainly effective security against fake news. The results are compared to other approaches thanks to a publicly available dataset, which shows a very satisfactory performance with a precision of 63.74%, an accuracy of 59.33%, an F1-score of 71.66% and Matthews Correlation Coefficient (MCC) with 56.61%. This study allows integrating social media technologies, and artificial intelligence to avoid fake news. The training is combined with educational awareness to always carefully retrieve safe pattern information from multiple social media sources while improving the CNN-LSTM-based alert model. Finally, the hybrid model is evaluated on the Coronavirus Disease 2019 (COVID-19) health crisis to obtain promising results compared to other approaches. This comparison shows extremely positive educational effects on reducing health crisis alerts in sustainability.