| China is in an important stage of economic structure and industrial structure adjustment and optimization.Small and micro enterprises(SMEs)play a key role in promoting the adjustment of C hina’s industrial structure and the sustainable and smooth operation of the social economy due to their diverse industrial types,wide coverage and sensitive market response.However,the financing difficulties of SMEs,especially the dilemma of traditional financing,limit the development of SMEs,and the turn to private credit financing makes SMEs face the risk of expensive financing and intensifies market turbulence.Information asymmetry is the key reason for the financing dilemma of SMEs.In the context of "Internet+",accelerating C hina’s financial supply-side reform and promoting the development of Internet finance not only provide conditions for the expansion of financing channels for SMEs,but also provide important means for obtaining information related to the credit level of MSME financing.Internet financing for small a nd micro enterprises in the context of big data has become a powerful way to solve the financing dilemma of small and micro enterprises;an in-depth and detailed research on the credit evaluation of Internet financing for small and micro enterprises based on big data technology is the key for this way to work.The key issues researched in this paper include:(1)analysis of evaluation mechanism.We study the characteristics of foreign micro and small enterprises and the successful experience of credit service practice,analyze the sources of credit risk of micro and small enterprises in Internet financing,analyze the similarities and differences between traditional and Internet channels of micro and small enterprises,and analyze the credit evaluation mechanism of micro and small enterprises in Internet financing in the context of big data.(2)Construction of evaluation system.Accelerate the research of the credit evaluation system for micro and small enterprises on the Internet,strengthen the effectiveness of the credit evaluation of micro and small enterprises on the Internet,and become an important measure for banks and other financial institutions to serve micro and small enterprises on the Internet and prevent and control the risk of micro and small enterprises on the Internet;(3)data model construction.With the ability of Internet technology and big data technology to obtain massive and diversified information,how to improve the credit evaluation model of small and micro enterprises financing and the accuracy of the model by using data mining and machine learning technology;(4)Evaluation results and application.Researching the mechanism of credit evaluation on the formation of trust relationship between financial institutions and micro and small enterprises in Internet financing is a key issue that requires in-depth analysis and research.Based on the analysis of the basic theoretical model,the article generates the following research results through the study of four major categories of problems.(1)The evaluation mechanism model of Internet financing for micro and small enterprises based on big data.In view of the current situation that the Internet financial products,services and financing processes of micro and small enterprises are not c lear and systematic,the analysis of the Internet financing channels of micro and small enterprises and the credit evaluation mechanism of big data is launched.C hina’s micro and small enterprises are an important force in the development of the national economy.Starting from the analysis of foreign micro and small enterprises’ characteristics and credit service practices,detailed statistical analysis and comparative analysis of C hina’s micro and small enterprises’ characteristics are realized,and C hina’s micro and small enterprises show significantly different characteristics in terms of overall situation such as geographical distribution and industry distribution,financial behavior,competitiveness,social relevance,as well as repayment ability and repayment willingness.Based on the case study,we propose four types of Internet financing models for SMEs,including the Internetization of traditional financing business,the financing model based on the Internet platform,the new model of "Internet+" finance,and listed financing.We propose four types of risk sources of micro and small enterprises’ Internet financing credit: information asymmetry risk,financial risk,relationship complexity risk,and vulnerability risk,analyze the similarities and differences between traditional and Internet channels of micro and small enterprises’ financing credit risk evaluation,and define a four-dimensional conceptual model of micro and small enterprises’ Internet financing credit evaluation.Examine the application mechanism of big data technology in the evaluation of micro and small enterprises’ Internet financing credit,and propose the basic framework of evaluation.(2)Internet financing evaluation system model for micro and small enterprises with "Internet+".We analyze the factors affecting the credit level of micro and small enterprises’ Internet financing,construct a special index framework for micro and small enterprises’ Internet financing credit evaluation,and form a reserve of specific and detailed evaluation indexes by making full use of the advantages of accessing massive information brought by the widening of micro and small enterprises’ financing channels under "Internet+".We define the hard and soft information reflecting the credit level of micro and small enterprises in Internet financing,and refine 81 specific factors affecting the credit level of micro and small enterprises in Internet financing from dual dimensions.Following that,90 specific evaluation indicators are refined and reserved from six dimensions,including information on business activities,financial behavior,basic information of enterprises,information on enterprise competitiveness,information on enterprise affiliation system and information on enterprise financing applicatio n statement,to construct the general framework of the evaluation index system of micro and small enterprises’ Internet financing credit,and clarify the index type,scoring method and data standardization method of each evaluation index.The combination o f the entropy value method,C RITIC method and variance chi-square test is used to assign the weighting method of micro and small enterprises’ Internet financing credit evaluation indexes,and to strengthen the ability of the index system to distinguish the performing customers and defaulting customers of micro and small enterprises’ Internet financing credit.(3)Data acquisition model of micro and small enterprises’ Internet financing based on topic mining.The information volume explosion and diversified data structure are the distinctive features of the evaluation of micro and small enterprises’ Internet financing in the context of big data,and the research on micro and small enterprises’ Internet financing credit evaluation needs to focus on the integration of hard and soft information multi-structure.Based on some soft information variables in the index framework of micro and small enterprises’ Internet financing credit,we construct an LDA theme mining model of soft information related to micro and small enterprises’ Internet financing credit,obtain 17 theme variables and the probability distribution of each theme corresponding to each sample,and realize the quantitative mining of soft information of micro and small enterprises’ Internet financing credit.(4)A neural network-based evaluation model for micro and small enterprises’ Internet financing.For the dichotomous problem of micro and small enterprises’ Internet financing credit evaluation,a BP neural network-based micro and small enterprises’ Internet financing credit evaluation model is constructed.The new third board of the Flush Quotation Center was selected as the data source to carry out the case study,and hard information,soft information,and hard information + soft information were incorporated into the BP neural network model to form three groups of comparison experiments.The results found that hard information can reflect the solvency and willingness to repay of micro and small enterprises’ Internet financing more intuitively,and the evaluation accuracy of micro and small enterprises’ Internet financing credit evaluation using hard information alone is higher than that using soft information alone;the accuracy of micro and small enterprises’ Internet financing credit evaluation incorporating soft information is higher than that of evaluation using only soft information or hard information,therefore,soft information is incorporated Therefore,incorporating both soft and hard information into the credit evaluation model of micro and small enterprises’ Internet financing is an important way to improve the accuracy of credit risk assessment by banks and other financial institutions.(5)Policy suggestions for the development of Internet financing for micro and small enterprises.Taking the rationalization of multiple data as a vein,the mechanism of credit evaluation on the formation of trust relationship between financial institutions and micro and small enterprises in the context of big data is studied,and strategies and management suggestions to promote the benign development of the trust system of Internet financing between micro and small enterprises and financial institutions are given.The processing paradigm of banks and other financial institutions for multiple information at different stages is clarified,and the way in which multiple information plays a role for micro and small enterprises is analyzed.It abstracts the Internet financing behavior of micro and small enterprises and the lending behavior of banks and other financial institutions into a game problem of forming the Internet financing trust relationship between micro and small enterprises and financial institutions,constructs an evolutionary game model,analyzes the strategy choices of the two in the process o f forming the Internet financing trust relationship,reasoning backwards the influence of relevant factors in credit evaluation on the strategy choices of the two,and then provides corresponding management suggestions for banks and other financial institutions.This paper provides management suggestions for banks and other financial institutions.The innovation points arising from this paper may include.(1)The general framework of the credit evaluation system of Internet financing for small and micro enterprises based on big data.Starting from the definition of the concept and scope of micro and small enterprises in C hina,we propose a general framework of the index system for the evaluation of micro and small enterprises’ Internet financing credit in the context of big data,and provide a paradigm reference for the subsequent research.The problem of insufficient access to structural hard information such as financial management data in the financing process of micro and small enterprises makes it diffic ult for banks and other financial institutions to make accurate risk assessment,resulting in serious "information asymmetry risk",which in turn makes it more difficult for micro and small enterprises to raise funds on the Internet.Soft information,such as unstructured textual information,and diversified data structure are the distinctive features of the evaluation of micro and small enterprises’ Internet financing in the context of big data,and the research on credit evaluation of micro and small enterprises’ Internet financing needs to focus on the integration of multiple structures of hard and soft information.In view of the characteristics of C hina’s micro and small enterprises,we consider the dual dimensions of hard and soft information generated during the operation of micro and small enterprises,refine a number of factors affecting the credit level of micro and small enterprises’ Internet financing,build a general framework for the index system of micro and small enterprises’ Internet financing credit evaluation,and determine specific evaluation indicators and index types and standardization methods.Big data technology can greatly enhance the availability of different types and sources of data,thus this data collection becomes possible to obtain more data with the assistance of big data technology,which also provides reference for other related studies.(2)The construction method of the credit evaluation model of micro and small enterprises’ Internet financing based on the fusion of multiple information.To study the way of acquiring and using unstructured soft information under big data technology.The explosion of soft information and diversified data structure are the distinctive features of the evaluation of micro and small enterprises’ Internet financing in the context of big data.Based on text mining technology,the quantitative method of non-structural soft information of micro and small enterprises’ Internet financing credit evaluation in the context of big data is proposed to realize the quantitative fusion of structural hard information and non-structural soft information of micro and small enterprises’ Internet financing credit evaluation in the context of big data.Based on the machine learning method and neural network model,we construct the credit evaluation model of micro and small enterprises’ Internet financing in the context of big data.Using web crawler technology,we study the multivariate data acquisition method for China’s micro and small enterprises’ Internet financing credit risk evaluation.A triple experiment is designed for the case study,using hard information dataset,soft information dataset,and hard information + soft information dataset,and the three types of data sets are incorporated into the evaluation model,and it is found that the use of different data types has a significant impact on the accuracy of the model.Hard information can more intuitively reflect the solvency and willingness to repay of micro and small enterprises’ Internet financing,and the accuracy of evaluation using hard information alone is higher compared with that using soft information alone for micro and small enterprises’ Internet financing credit evaluation;the accuracy of micro and small enterprises’ Internet financing credit eva luation incorporating soft information is higher than that of evaluation using only soft or hard information.(3)The relationship model between financial institutions and micro and small enterprises’ trust in Internet financing based on evolutionary game.In view of the lack of mutual trust between micro and small enterprises and banks and other financial institutions in C hina,we construct an evolutionary game model of the trust relationship between financial institutions and micro and small enterprises in Internet financing,and analyze how to achieve the ultimate goal of improving credit evaluation and promoting mutual benefit and trust between micro and small enterprises and financial institutions in Internet financing through different strategic choices of Internet financing credit evaluation under big data technology.Based on the rationalization of multi-data,we abstract the Internet financing behavior of micro and small enterprises and the lending behavior of financial institutions into the game problem of forming the Internet financing trust relationship between micro and small enterprises and financial institutions,and build the evolutionary game model.Through the analysis of evolutionary game equilibrium and the analysis of influencing factors,we simulate the strategy choice of the two in the process of forming the trust relationship of Internet financing,and find that the potential gain and loss of financial institutions in the Internet financing of micro and small enterprises,the potential gain and loss of structural soft information reflecting the credit level of micro and small enterprises,and the mastery of soft information hardening boundary,machine evaluation and interpersonal communication relationship are important factors affecting the strategy choice of the two and the formation of good trust relationship.important factors in the formation of good trust relationship.Based on this,corresponding management suggestions are provided for banks and other financial institutions in order to realize the improvement of credit evaluation effect of Internet financing for micro and small enterprises in the context of big data.There are 58 figures,27 tables and 204 references in this thesis. |
PDF Full Text Request