In many applications, data mining techniques are used as a regular practice to analyze the vast amount of available data and extract relevant knowledge and information to support the main decision-making processes. The content-based document classification system assigns a document to one of the specified classes by using the content and some weighting criteria. The classification of documents using Random Forest, K-Nearest Neighbor, Multinominal Naive Bayes, Multinominal Logistic Regression, and Support Vector Machine is examined in this study. Here, the metadata parameters were chosen from the seven subjects collected from the IEEE dataset domain, such as title, author keywords, and IEEE terms. It is used for classifying data into different classes by considering some constraints. In order to give the best outcome, we compare these five algorithms. The Logistic regression classification technique performs better which might also aid a Course Recommender System