In recent times, sentiment analysis is one of the most important aspects of machine learning research. However, many context-aware systems have developed based on the English language, which can automatically process the English language to make a clear emotion. However, much less work has done on Bengali than that. The main reason for this is the lack of an accurate Bengali dataset. In our works, we have used a unique dataset. The data is mainly various comments made by people through online news portals and social media. Humans collect those and maximum awareness sought while labeling the emotions. All data labeled into positive (1) and negative (0) emotions. Our main objective in this research is to build a Bengali context-aware system using various supervised machine-learning algorithms that can easily find out the emotions of any Bengali language. For this, we used K-Nearest Neighbors (KNN), Decision Tree (DT), Logistic Regression (LR), Support Vector Classifier (SVC), Multinomial Naïve Bayes (MNB), and Random Forest (RF) algorithm. Among them, the Random Forest (RF) algorithm generates the maximum accuracy that is 67.34 %.