Evaluation method of e-government audit information based on big data analysis

In order to improve the problems of low efficiency of intelligent auditing and long approval cycles in the evaluation of e-government audit information, this paper proposes a model for the evaluation of e-government audit information. The model covers the acquisition of audit information, evaluation rules, and e-government auditing. The intuitionistic fuzzy roughness model is applied to determine the distinguishing relationship between objects based on the intuitionistic fuzzy values of different objects and to determine the evaluation rules for audit information. An e-government auditing approach based on big data analysis is also proposed, and a genetic algorithm is used to optimize the relevant parameters of a convolutional neural network to enhance the stability of the e-government auditing model. Audit information evaluation activities are carried out at Changchun Housing Provident Fund Center in Jilin Province, China, whose shared approval information data failure records in 2023 accounted for as high as 70.54% of all documents, with low stability. The number of cases and closure of the provident fund center accepting the extranet from 2017 to 2023 only accounted for 4.57% and 3.99% of the whole. In the audit information evaluation indicators, nine secondary indicators, such as credit management and indirect income, have an evaluation value of more than 0.5 in the “medium” evaluation level, and many indicators have an evaluation value of more than 0 in the “poor” evaluation level, so there are many shortcomings in the business.