Identify Pragmatic Functions of Students” Analytical Texts via Feature-Based Machine Learning Method
Providing instant and reliable feedback for students” writings over a long project span is a time-consuming and labor-intensive task.Many studies have implemented automatic approaches to relieve the burden from teachers.However,traditional approaches only applied limited evidence of linguistic phenomena,and no study has based the task on the scenario of students” analytical report in Chinese.Targeting Chinese texts of students” analytical report,this study implemented machine learning models with a comprehensive set of features for differentiating pragmatic functions of texts.This paper reports the processes of corpus building,annotation,feature extraction,model training,and evaluation.Results indicate the effectiveness of the classification model improved by a feature selection procedure.
Chinese text classification feature-based machine learning pragmatic role identification
Chen Qiao Xiao Hu
Division of Information & Technology Studies,The University of Hong Kong
国内会议
北京
英文
546-549
2017-06-02（万方平台首次上网日期，不代表论文的发表时间）