AUTOMATIC CLASSIFICATION OF EFL LEARNERS’ SELF-REPORTED TEXT DOCUMENTS ALONG AN AFFECTIVE CONTINUUM

This study aims to place EFL learners along an affective continuum via machine learning methods and present a new dataset about characteristics of learners. In line with the purposes, written self-reports 475 students from 5 different faculties in 3 universities Turkey were collected manually assigned by researchers one labels (positive, negative, or neutral). As result, two combinations same (AC-2 AC-3) including numbers classes used for assessment automatic classification approaches. Results revealed that confirmed manual great extent could be classify according their characteristics. Maximum accuracy rate is 90.06% on AC-2 classes. Similarly, AC-3 three classes, maximum 71.79%. Last, top-10 features/words obtained feature selection are highly discriminative terms assessing student feelings learning. It stated there not existing which classifiers literature automatically learners’ feelings.