This paper presents an experimental study of supervised algorithms for text classification. Naive Bayes, Support Vector Machines, Random Forests, Decision Trees, KNNs, Neural Networks, and Logistic Regression have been compared. These algorithms are tested and compared on the Weka tool. For this experiment, a dataset of two thousand sentences with parts of speech ambiguity has been collected. The collected data are organized and pre-processed by removing stop words and feature extraction. The results of the comparison are based on the F-score, recall, and precision values returned by each algorithm. Results show that out of these seven classifiers, Decision Tree is computationally efficient and shows a higher accuracy percentage. To enhance the accuracy of the classified document, we have proposed a hybrid model. In this model, we have integrated the SVM, Decision Tree, and Naive Bayes’ algorithm to get a more accurate result as compared to Decision Tree. This classification approach is coined “AmbiF”. The accuracy of all analyzed algorithms ranges between sixty-six to eighty- four percent while for AmbiF model it is reported as eighty-five percent.