Machine Learning-Driven Implicit Evaluation in College English Ideological Education: A Hybrid Corpus-Linguistic and Survey-Based Approach

In this paper, discourses in English have been analyzed by means of UAM Corpus Tool, in the light of Appraisal Theory. It counts the distribution of engagement resources and investigates their contributions to expressions of evaluation and intersubjectivity. The questionnaire survey gathers the feedbacks on implicit evaluation from 310 students in the College English Course. It processes the survey data by machine learning, in particular support vector machines (SVM). It presents quantitative perspective on the opinions of students toward the implicit evaluation. The adopted SVM model raises accuracy of students’ opinion, and it reveals hidden patterns in students’ opinions. The study indicates that implicit evaluation of the engagement system serves College English teachers. It makes their lectures more persuasive and promotes students’ critical thinking ability. It also guides students’ values and enhances their ability of international communication.