Application of Social Network Analysis and Autoencoder in Enterprise Internal Collaboration Optimization System

It is difficult for the existing workflow modeling technology to effectively deal with the complex and dynamic characteristics of business processes between enterprises, especially to support the ability of process collaboration. This project intends to study the modeling theory and method of flexible workflow system based on business agent from the perspective of multi-agent cooperation and high flexibility requirements. Construct a flexible workflow architecture with multi-level and multi-authority control. In this paper, the process management and implementation of hierarchical management and control. The process is optimized using social network analysis and self-encoder method. The advantage is that the original input can be restored in the case of noise, while taking full advantage of the diversity and complexity of noise. The deep neural network is used to enhance the recovery ability of the initial features, so as to effectively suppress the external noise. Through a case study of the actual process of a cashmere enterprise, the corresponding model is established. The research proves that the model proposed in this paper can deal with the cooperative relationship between process flexibility and enterprise organization well.