Public security of network big data from the perspective of entrepreneurship

Abstract

This study focuses on enhancing network big data public security governance through an entrepreneurial perspective. It begins by analyzing the current state of network big data public security governance and its theoretical foundations in entrepreneurship. The study then introduces an innovative network big data governance model tailored for urban grassroots society. Utilizing a meta-object mechanism, a social network big data public security governance scenario model is developed. The model’s performance is evaluated using gene structure representation and transfer learning methods, supported by real-world examples. Results indicate substantial regional disparities in entrepreneurship levels across Chinese provinces, with higher concentrations in the eastern region. Additionally, entrepreneurship positively influences the development of network big data, highlighting regional variations in its evolution. The spirit of innovation has a positive correlation with the development level of big data, and the fitting coefficient is 13.3. The average classification accuracy rate of gridded events in the study area is 85%, and the classification accuracy rate of four types of events is less than 80%. After the introduction of cross-region transfer learning, the classification effect of seven out of ten types of events has been improved, and the number of event classes with an accuracy less than 80% has been reduced from 4 to 1. The classification accuracy of “illegal buildings” increased by 6%, and the overall classification accuracy increased by 2.4%. The constructed model can improve the level of public safety governance. These conclusions provide references for the application of entrepreneurship in the complex system processing.