Features selection of high quality essays in automated essay scoring system

In this paper, we analyze some essays of Chinese English learners, and text features such as words, phrases, paragraphs, and chapters. According to the correlation between the scores and these features, we do multiple regression analysis based on the feature recombination and extract the feature sets that high quality essays should be with. In the end, we build a reasonable model to distinguish high quality essays from others and effectively solve the problem of low differentiation scores in automated essay scoring system.