The evaluation of teaching reform satisfaction is crucial for improving teaching quality. In the evaluation of satisfaction with teaching reforms in environmental art and design, issues such as the low reliability of evaluation indicator data can lead to unsatisfactory results. To address this, this paper proposes a new method for evaluating satisfaction with teaching reforms in environmental art and design based on multi-source data integration. By establishing principles for constructing a multi-source indicator system for evaluating satisfaction with teaching reforms in environmental art and design, a four-level multi-source indicator system was determined. This system mainly includes one first-level indicator, four second-level satisfaction evaluation indicators, and several third- and fourth-level evaluation indicators. The construction of the multi-source indicator system for evaluating satisfaction with teaching reforms in environmental art and design was completed. Using a four-part graph model and entropy weight calculation, the importance of the evaluation indicators for satisfaction with teaching reforms in environmental art and design was classified. The expert scoring method was introduced to correct the multi-source indicator data for evaluating satisfaction with teaching reforms in environmental art and design. Based on standardized indicator data, a penalty rule was added to the loss function of the multiple linear regression model to remove data noise, achieving pre-processing of the multi-source indicators for evaluating satisfaction with teaching reforms in environmental art and design. After completing the above process, first collect multi-source data from questionnaire surveys, student feedback, etc., and preprocess and map it to a unified ontology class. Subsequently, data features are extracted through non-negative matrix factorization, and the similarity between object types is calculated. Finally, based on the importance, redundancy, and interaction of data sources, an integrated quality model is designed to effectively integrate multiple sources of data, construct a satisfaction evaluation model, and output the score of satisfaction evaluation. The experimental results show that the proposed method has the highest satisfaction index, reaching about 0.99, indicating a high degree of recognition for the reform; when the amount of integrated data changes, the proposed method maintains the best integratison reliability. Even with an increase in data volume, its reliability only slightly decreases and stabilizes at a high level; meanwhile, the average absolute error of the proposed method is the lowest, consistently below 2.5%, indicating a high level of stability and accuracy in its evaluation results. The above results confirm that the proposed method can effectively evaluate the satisfaction of environmental art design teaching reform and improve teaching quality.
Read full abstract