For the people of the 21st century, COVID-19 is a name of fear. So many people have surrendered to this disease before realizing anything. After so much has happened, many people still take it less seriously and don’t follow the standard hygiene rules given by the WHO. The purpose of this research work was to predict health awareness among the people of Bangladesh in post-COVID-19 era. To perform this work, we needed a huge amount of data. For this reason, we conducted a survey through online and offline medium. Eventually, we were able to collect 1,023 data from two divisions (Dhaka and Khulna) in a very small range. The amount of data is too less. But we were not able to collect more than this because of some unwanted problems. So, ultimately, we had to proceed with this much number of data. We used 20% data for testing and the rest 80% of data for training the models. We applied 6 different types of machine learning algorithms such as Logistic Regression, Decision Tree Classifier, KNN (K-Nearest Neighbor), SVM (Support Vector Machine), Random Forest Classifier and Gradient Boosting Classifier. Random Forest Classifier outperformed all the other machine learning algorithms achieving the highest accuracy of 96.58%, highest F1 score of 96.42%, highest precision of 95.83% and highest recall of 97.26%.

COVID-19 Health awareness Survey Machine Learning Logistic Regression Decision Tree Classifier KNN (K Nearest Neighbors) Random Forest Classifier SVM (Support Vector Machine) Gradient Boosting Classifier.