Natural Language Processing And The Research Of The Risk Management Of Folks Lending Of Small Enterprises

The shortage of operating capital for small and micro enterprises has always existed.China's formal financial institutions cannot meet their financing needs.Under this circumstance,the phenomenon of providing small financing services to small and micro enterprises or individuals in the form of folks loan has appeared in large numbers,becoming an important phenomenon in the field.The transparency of folks loan is relatively low,and the lending behavior is not normative and professional.As a result,the issue of folks loan by small and micro enterprises is frequent,and various problems are reflected in the court proceedings.The main research work of this paper is to analyze the risk factors of private lending in small and micro enterprises,apply Natural Language Processing technology to analyze litigation cases,and give solutions and policy recommendations for folks loan risk management of small and micro enterprises.First of all,based on relevant laws and regulations,small and micro enterprise financing and other related literature,this paper divides risk factors into macro and micro levels.Through interviews with 25 senior judges in Zhejiang Province,the problems of macro-level laws and regulations,lending strategies,and professional ability of capital supply subjects were sorted out.The 3W model was used to sort out the risk factors of micro-level private lending.The game analysis of small and micro enterprise financing is carried out to construct two sets of borrowing and lending game matrices in the economic upward(including equilibrium)cycle and the economic down cycle.Through the operation of the game matrix,the condition for achieving the maximum return equilibrium state of both parties is to do credit access management,credit investigation management,credit enhancement management,and new technology applications.Secondly,in the judgment of folks loan cases involving small and micro enterprises,the lending process,borrowing factors and controversial issues are recorded in detail.The use of natural language processing technology(NLP)to construct text classification algorithms can identify risk factors.Considering that the judgment is logically strong,a viewpoint is usually expressed by multiple clauses.For this continuous text sequence,a text classification algorithm is trained based on a deep learning model-recycling neural network(RNN)that can automatically extract features.With K-fold cross-validation,algorithm training and verification have achieved good results.Using the well-trained algorithm,the risk factors in the judgment were identified,and various risk factors were statistically analyzed,and the path of the folks loan risk management system was found.Thirdly,after combined with the experience of formal financial risk prevention and control,I believe that: First,we must further regulate private lending from the legal level,clarify the legal status of private lending,clarify the regulatory body,and crack down on illegal Private lending,to prevent market disorder caused by the blank of law,to provide protection for SMEs through private lending financing.Second,the characteristics of intermediation and professionalization of private lending are becoming more and more obvious,and risk management has a high professional threshold.The supervisory department should refer to the mature experienceindomestic and foreign to promote the private lending and lending system.Under the loan law,the private lending practitioners or organizations that have reached the licensing standards are given entry qualifications,and the operating conditions of the licensed practitioners or organizations are assessed,and thosepractitioners or organizations that fail to meet the standards are required will exit.Third,the lack of credit investigation is the most common risk factor among private lending risks.By improving the credit reporting system of SMEs,open up more data sources such as finance,taxation,industry and commerce,water and electricity,and incorporate more business data from SMEs.It can effectively help lenders identify risks and lower the barriers to loan investigation.Fourth,the regulatory authorities should promote the establishment of a market-based private lending service center,establish an electronic service platform with information services,payment and settlement,and professional credit services as the core capabilities,and provide professional,low-cost,private lending through social division of labor.