Prediction of Essay Scores From Writing Process and Product Features Using Data Mining Methods

Abstract

Analysis of keystroke logging data is of increasing interest, as evident from a substantial amount of recent research on the topic. Some of the research on keystroke logging data has focused on the prediction of essay scores from keystroke logging features, but linear regression is the only prediction method that has been used in this research. Data mining methods such as boosting and random forests have been found to improve over traditional prediction methods such as linear regression in various scientific fields, but have not been used in the prediction of essay scores from keystroke logging features. This article first provides a review of boosting, which is a popular data mining method. The article then applies boosting to predict essay scores from a large number of keystroke logging features and other predictor variables from two real data sets.