Combine DGBL with AI system: A Technical Guidance to Reduce Teacher’ s Burden in Digital Game-based Learning


  • yue lei
  • liang guo



Game-based learning has been regarded as a increasing popular method in current teaching process, however, it really burdens teacher as it requires teachers to invest abundant energy and time to design. Aims at reducing teaching burden, this research has proposed a guidance system design based on large language model (LLM) and blockchain technology. In this design, system framework has been divided into 3 layers: user layer, application layer and technical layer. Initially, teachers input their instructional plans, while students signing up their learner profiles. This information is securely recorded on the blockchain for data integrity. The results stemming from data prediction and feature engineering are then incorporated into the LLM , facilitating the visualization of strategies tailored to address specific learning challenges. As the process advances, the information undergoes automated scrutiny to evaluate the learning conditions, ultimately selecting an appropriate DGBL cases with a proven track record in similar scenarios. This aids teachers in crafting personalized learning blueprints, informed by the insights gleaned from the feature engineering analysis and its impact on students' learning experiences. The concluding phase involves tracking and assessment, wherein an automated evaluation of student performance is conducted based on study data and LLM-generated questionnaires. Teachers subsequently review the results and recommendations to enhance the quality of their instructional methodologies, and the learner portrait will also be renewed according to received data. This guidance system still has some disadvantages, such as lacking sequential consistency in the responses generated by the model. In summary, a future direction for this research is to develop specific LLM systems for specific school segments and instructional needs to help teachers implement DGBL