Descriptive Answer Script Grading System using CNN-BiLSTM Network

Descriptive answer script assessment and rating program is an automated framework to evaluate the scripts correctly. There are several classification schemes in which a piece of text evaluated on basis spelling, semantics meaning. But, lots these aren’t successful. Some models available rate response include Simple Long Short Term Memory (LSTM), Deep LSTM. In addition that Convolution Neural Network Bi-directional LSTM considered here refine result. The model uses convolutional neural networks bidirectional learn local information words capture long-term dependency contexts Tensorflow Keras deep learning framework. embedding semantic representation texts can be used for computing similarities between pieces grade them based similarity score. experiment methods data optimization, such as normalization dropout, tested Automated Student Evaluation Response Scoring, commonly public dataset. By comparing with existing systems, proposed has achieved state-of-the-art performance achieves better results accuracy test