| In traditional tests,teachers often pay attention to students’ scores,which directly reflect students’ learning.However,it is coarse in granularity,and it is difficult to provide a detailed description of students’ cognitive structure,and it is impossible to reflect the specific knowledge points of students,which is not conducive to the improvement of students’ knowledge level and hinders the improvement of teaching quality.Knowledge tracing focuses on students’ long-term mastery of knowledge,and obtains students’ knowledge mastery through analyzing historical learning data,which can help teachers grasp important and difficult points,thus adjust teaching strategies,and then improve teaching effectiveness.Based on the standard Bayesian knowledge tracing model,this study proposes a new knowledge tracing model,the DT-BKT model,by introducing an item difficulty parameter and setting a more reasonable initial state transfer probability to address the shortcomings of the standard Bayesian knowledge tracing model,which does not take into account the characteristics of items and the initial parameter setting.At the same time,the DT-BKT model is compared with the benchmark model using a publicly available dataset to evaluate its performance.The experimental results show that the DT-BKT model outperforms the benchmark model in terms of AUC,RMSE,accuracy rate and loss rate.Based on this,this study applied the DT-BKT model to analyze the test data of a senior high school information technology course to obtain the students’ mastery of each knowledge point.At the same time,class diagnostic reports and individual student diagnostic reports were generated based on the results of the model analysis to help teachers tailor their teaching and motivate students to target and fill in gaps.Finally,interviews with teachers and students were conducted to verify the feasibility and effectiveness of applying the improved model to senior secondary information technology courses for teaching evaluation. |