Data Mining of the Underachievers’ Performance of E-learning and Finals in College English with SPSS and WEKA

This study aims to explore the academic performance of underachievers in college English through the perspectives of e-learning engagement and learning outcomes by using the statistical means of SPSS and the predictive modeling of J48 Decision Tree with WEKA. The statistics in iTEST, paper-based and computerized final exams by underachievers were analyzed with regard to the differences in genders and majors and the correlations mainly through SPSS. The J48 Decision Tree in WEKA provided a nuanced understanding of the hierarchical nature in learning factors and their effect on success in the final exam. The integration of SPASS and WEKA made it possible to generate feasible approaches for pedagogical strategies, and personalized learning instructions to enhance the underachieving students’ academic achievement in the e-learning environment.