Prediction Method of Online Teaching Effect of Career Guidance Course Based on Multi Task Learning

There are issues with intermittent and lagging schedules in the teaching of employment guidance courses, as well as a lack of professional teachers. Online teaching can solve these problems. Therefore, in order to better analyze the learning effectiveness of online teaching in employment guidance courses, a multi task learning based prediction method for the learning effectiveness of online teaching in employment guidance courses is designed to determine whether learners have mastered the knowledge content of the employment guidance courses they are learning. Design a Chinese word segmentation sequence annotation method based on multitasking learning, implement preprocessing of online teaching data for employment guidance courses, and improve the generalization ability and effectiveness of the model. For the data after Chinese word segmentation, an improved K-means algorithm is used to mine the learning behavior information of students, thereby discovering potential patterns and patterns hidden in the data. Propose the EduHawkes model, which adds various types of feature information to predict learning outcomes based on analyzing student behavior patterns, achieving online teaching of employment guidance courses and improving the accuracy and interpretability of learning outcomes prediction. The test method shows that the accuracy of the method and the Matthews correlation coefficient are both high.