Research On Key Technologies Of Industry Identification And Credit Portrait Of Private Enterprises In Sichuan Province Based On Big Data

As an important achievement in the development of China’s socialist market economy,the private economy has become an important force in promoting the development of China’s socialist market economy.The database of the Sichuan Standardization Research Institute stores the data of all private enterprises in the province.However,due to the business scope data of private enterprises in this database being manually inputted in the early collection,there are real problems such as unclear,non-standardized,and imperfect business scope data,resulting in the inability to effectively determine the industry categories these private enterprises are engaged in.At the same time,the Sichuan private economy comprehensive service platform lacks a credit assessment model for enterprises,which affects the efficiency of its services for private enterprises.In view of the above problems,this thesis focuses on the data related to private enterprises to carry out the key technology research of private enterprise industry identification and credit portrait in Sichuan Province,the main work is as follows:1.Identifying the categories of industries in which private enterprises are engaged employing word similarity calculations and building multiple machine learning models;By comparing the advantages and disadvantages of these methods,an innovative method integrating text similarity calculation technology and machine learning models is designed and implemented.This method semi-automatically annotates enterprise data by calculating text similarity and uses the annotated data to perform multi-label supervised training of machine learning models to identify the industry categories of enterprises,which can solve the time-consuming problem of identifying the industry categories of enterprises while maintaining good accuracy.2.Based on the relevant data of private enterprises,the analytic hierarchy process is used to make credit portraits of enterprises,and the rationality of the credit portrait model is verified by substituting enterprise data.The credit score calculated by this work can directly reflect the credit status of enterprises,and provide an effective reference for private enterprise credit and government financial support for private enterprises.3.In addition,this thesis makes a statistical analysis of the industry and region data of the enterprises in the province.This work can help users understand the development trend of enterprises in the province so that the platform can serve enterprises in a more targeted manner.