The issue of incorrect repertoire selection arises despite the importance of model development in the process of vocal art singing repertoire selection. The model creation issue in vocal art singing repertoire selection is intractable and produces unsatisfactory results when using the conventional decision tree technique. As a result, this work suggests and examines the development of a model for selecting vocal art singing repertoires using the Plain Bayes algorithm. To begin, we utilize Bayes' theorem to identify the elements that will have an impact, and then we partition the indicators based on the needs of the model building process to lessen the impact of any interfering factors. The model creation results are then thoroughly examined after a naïve Bayesian algorithm is created using Bayes' theorem. In terms of model building influencing factor time and model construction accuracy, the MATLAB simulation results demonstrate that the Naive Bayes method outperforms the classic decision tree approach under certain assessment criteria.