ENGLISH DISTANCE TEACHING BASED ON SPOC CLASSROOM AND ONLINE MIXED TEACHING MODE

At present, the English Small Private Online Course (SPOC) online mixed teaching model has problems in evaluating students' learning and organizing teaching papers. For example, the evaluation is chaotic and unable to meet the key points of organizing the paper. Starting from the thinking chain that accepting learning outcomes can promote learning behavior, a score prediction method and test paper generation algorithm (TPGA) based on a learning evaluation diagnostic model are designed. Among them, the performance prediction algorithm is designed by combining multiple linear regression (MLR) and random forest (RF). The TPGA is based on students' learning status. The research results show that most of the predicted values output by the performance prediction model are not significantly different from the actual values. They are within a reasonable range. Meanwhile, under the influence of TPGA, the number of students in the experimental group is higher in the 70-80 and 80-90 segments, with 27 and 6, respectively. The experimental group has a higher average score rate on each type of question and knowledge point. Both models have high student satisfaction, indicating that the results oriented online mixed learning strategy designed in the study can effectively improve students' learning outcomes.