With the new computational techniques, one can use sentence comprehension and sentence classification with more ease and convenience. This study explores methods of making representations of sentences meaningful using computers. Various methods are proposed for modeling sentence structure and meaning by representing sentence data in a new way for classification. The study outlines a number of ways for generating these models, ranging from random to tree-structured and evaluates them. The results indicate that these models are highly accurate as different ways produce similar performance for both training and testing. Problems found while creating the model are discussed along with what can be done about it in the future.