Search

To improve students’ learning performance through review learning activities, we developed a personalized intervention tutoring approach that leverages learning analysis based on artificial intelligence. The proposed intervention first uses text-processing artificial intelligence technologies, namely bidirectional encoder representations from transformers and generative pretrained transformer-2, to automatically generate Python programming remedial materials; subsequently, learning performance prediction models constructed using various machine learning methods are used to determine students’ learning type, enabling the automatic generation of personalized remedial materials. The participants in this study were 78 students from a university in northern Taiwan enrolled in an 8-week Python course. Students in the experimental (n = 36) and control (n = 42) groups engaged in the same programming learning activities during the first 5 weeks of the course, and they completed a pretest of programming knowledge in Week 6. For the review activity in Week 7, the 36 students in the experimental group received personalized intervention, whereas those in the control group received traditional class tutoring. We examined the effect of the self-regulated learning and personalized intervention on the learning performance of students. Compared with the traditional class tutoring, the personalized intervention review activity not only helped students obtain higher learning performance but also prompted greater improvements in the following learning strategies: rehearsal, critical thinking, metacognitive self-regulation, effort regulation, and peer learning. We also observed that students’ rehearsal and help-seeking learning strategies indirectly affected learning performance through students’ note-taking in the provided e-book.

To improve students' learning performance through review learning activities, we developed a personalized intervention tutoring approach that leverages learning analysis based on artificial intelligence. The proposed intervention first uses text-processing artificial intelligence technologies, namely bidirectional encoder representations from transformers and generative pretrained transformer-2, to automatically generate Python programming remedial materials; subsequently, learning performance prediction models constructed using various machine learning methods are used to determine students' learning type, enabling the automatic generation of personalized remedial materials. The participants in this study were 78 students from a university in northern Taiwan enrolled in an 8-week Python course. Students in the experimental (n = 36) and control (n = 42) groups engaged in the same programming learning activities during the first 5 weeks of the course, and they completed a pretest of programming knowledge in Week 6. For the review activity in Week 7, the 36 students in the experimental group received personalized intervention, whereas those in the control group received traditional class tutoring. We examined the effect of the self-regulated learning and personalized intervention on the learning performance of students. Compared with the traditional class tutoring, the personalized intervention review activity not only helped students obtain higher learning performance but also prompted greater improvements in the following learning strategies: rehearsal, critical thinking, metacognitive self-regulation, effort regulation, and peer learning. We also observed that students' rehearsal and help-seeking learning strategies indirectly affected learning performance through students' note-taking in the provided e-book.

To improve students' learning performance through review learning activities, we developed a personalized intervention tutoring approach that leverages learning analysis based on artificial intelligence. The proposed intervention first uses text-processing artificial intelligence technologies, namely bidirectional encoder representations from transformers and generative pretrained transformer-2, to automatically generate Python programming remedial materials; subsequently, learning performance prediction models constructed using various machine learning methods are used to determine students' learning type, enabling the automatic generation of personalized remedial materials. The participants in this study were 78 students from a university in northern Taiwan enrolled in an 8-week Python course. Students in the experimental (n = 36) and control (n = 42) groups engaged in the same programming learning activities during the first 5 weeks of the course, and they completed a pretest of programming knowledge in Week 6. For the review activity in Week 7, the 36 students in the experimental group received personalized intervention, whereas those in the control group received traditional class tutoring. We examined the effect of the self-regulated learning and personalized intervention on the learning performance of students. Compared with the traditional class tutoring, the personalized intervention review activity not only helped students obtain higher learning performance but also prompted greater improvements in the following learning strategies: rehearsal, critical thinking, metacognitive self-regulation, effort regulation, and peer learning. We also observed that students' rehearsal and help-seeking learning strategies indirectly affected learning performance through students' note-taking in the provided e-book. [ABSTRACT FROM AUTHOR]