A Virtual Simulation Technology based Decision Trees for Quality Evaluation of Ideological and Political Education in Universities

Nowadays, the evaluation of education quality has become a pressing concern globally, driven by the need to ensure that educational institutions provide students with knowledge which are required to succeed in an increasingly complex and interconnected world. The existing methods for quality evaluation of political and ideological education in universities has some limitations like limited assessment scope, lack of personalization. Therefore, this paper proposes Virtual Simulation Technology- Decision Tress for quality evaluation of Ideological and Political Education (IPE) in universities. Initially, the data is taken from EdX dataset from Harvard university. Then, this text data is preprocessed by using Natural Language Processing-Collaborative Filtering (NLP-CF) which involved text tokenization, stopword removal, lemmatization and in collaborative filtering data normalization is used. After that, feature selection is done by using Recursive Feature Elimination (RFE) which effectively selected the relevant features which are preprocessed. Finally, the quality is evaluated by using proposed Virtual Simulation Technology-Decision Tress which effectively enhanced the personalization by adapting individual learner profiles and identified key factors in IPE which influenced the learning outcomes. The proposed Virtual Simulation Technology- Decision Tress achieved better results in average error of $\mathbf{9 0. 4 3 \%}$, MPE of $\mathbf{9 2. 7 8 \%}$ when compared with existing RBFNN.