An Automated Essay Scoring model Based on Stacking Method

In the latest two decades, thanks to AI technology, Automated Essay Scoring (AES) technology has also been rapidly developed. This technology to analyze and score essays automatically is a hot spot for the application of natural language processing in the field of education. Firstly, different encoding methods are used to obtain the lexical vectors of the text. Secondly, features of the seven aspects of the text are fully extracted. Then the comparative analysis of different ensemble learning models was studied on the English essay scoring competition dataset of Kaggle. Finally, a model based on the stacking method is proposed. The results show that the model based on the stacking method achieves the best results on all datasets. It improves the average QWK values on eight subsets by 1.0%similar to 2.8% compared to the baseline models of neural networks and feature-engineered.