Smart Grading System Using Bi LSTM with Attention Mechanism

Essay assignment is one of the methods that help teachers to understand the writing skills of the students. Often these assignments are graded manually. Courses enrolled with hundreds of students put a lot of pressure on professors as delivering individual grades is a repetitive and time-consuming task. This project focuses on identifying the algorithm that will provide a computer-assisted model for grading of essays. In this project when an essay is loaded into the proposed grading system, the system accepts the essay given as the input and grades it using deep learning techniques. Proposed paper consists of a comparative study of models with different combinations of neural network algorithms and its layers. The models are compared on the basis of Quadratic Weighted Kappa (QWK) as evaluation metrics.