XiuminChen. Automatic Scoring of English Composition Based on Word Vector Clustering and Random Forest. Dynamic Systems and Applications 29 (2020) No. 5, 2090 – 2099

https://doi.org/10.46719/dsa202029543

ABSTRACT.
The purpose of this study is to use an automated scoring technique to assess second language composition in English. It aims at automatic scoring for classifying oral performance data of a large group of students into a small number of discrete oral proficiency levels. To use different natural language processing techniques and huge natural language tools, take a look at the different patterns that appear in the composition to get them. It can, after extracting the required formations from the creation of the dataset, further the composition given by the user and also make suggestions for those composition with obvious characteristics. After extracting the features that predict the score that through the training of a Word Vector Clustering Based Random Forest (WVCRF) algorithm that computes the cosine distance and recommends those composition to the user’s machine learning agent to determine similar informative composition.  It have also developed our system to show our mistakes in writing style and the necessary corrections for the writer and automatic scoring to English composition.

Keywords: Word Vector Clustering Based Random Forest (WVCRF), English Composition.