Integrated Comprehensive Analysis Method for Education Quality with Federated Learning

A thorough evaluation of education quality can help schools, governments, and related organizations evaluate education quality, detect persistent problems, act quickly, and raise teaching and educational standards. Traditional methods of assessing education quality, such as statistical analysis and teacher evaluations, suffer from limitations such as data fragmentation, privacy preservation and relying solely on one form of evaluation. We presented a method for comprehensively assessing education quality by incorporating federated learning (CEQFL) to fully utilize big data. First, under a unified evaluation model and indicators, universities act as terminal nodes to train the evaluation model using private data locally. Second, universities aggregate the model to the Ministry of Education for model aggregation and update the evaluation model parameters based on the aggregated data. Finally, using school data as experimental data, we demonstrate that CEQFL has strong education quality assessment capabilities.