CN-121998428-A - Entrepreneur risk assessment method based on industry data analysis

CN121998428ACN 121998428 ACN121998428 ACN 121998428ACN-121998428-A

Abstract

The invention discloses a startup risk assessment method based on industry data analysis, which belongs to the technical field of startup risk calculation and comprises the following steps of firstly setting a double-layer assessment module and a summarization module, wherein the double-layer assessment module comprises a preliminary field module and a subdivision assessment module, secondly collecting industry data in different fields to form a startup risk field library, and a data supply module is arranged in the startup risk field library, thirdly, inputting startup project related description content by a user, combining the preliminary field module with the data supply module, confirming corresponding startup data, classifying the acquired startup data, fourthly, inputting the classified startup data to the subdivision assessment module to obtain a preliminary assessment result, and fifthly, the summarization module obtains total startup risk according to the preliminary assessment result. By adopting the method, the invention carries out continuous quantitative calculation on the existing startup risk and carries out long-acting risk assessment.

Inventors

CAO ZHANXIA
YANG LI
HAN MIN

Assignees

郑州职业技术学院

Dates

Publication Date: 20260508
Application Date: 20260127

Claims (8)

1. A startup risk assessment method based on industry data analysis is characterized by comprising the following steps: setting a double-layer evaluation module and a summarization module, wherein the double-layer evaluation module comprises a preliminary field module and a subdivision evaluation module; acquiring industry data in different fields to form an entrepreneur risk field library, wherein a data supply module is arranged in the entrepreneur risk field library, and the data supply module normally updates the content of the entrepreneur risk field library according to the set updating time; step three, a user inputs entrepreneur related description content, a preliminary field module is combined with a data supply module, a specific field corresponding to the entrepreneur related description content in an entrepreneur risk field library is confirmed based on the entrepreneur related description content, the entrepreneur risk field library correspondingly outputs entrepreneur data of the specific field, and the data supply module classifies the acquired entrepreneur data; Inputting the classified startup data into a subdivision evaluation module to obtain a preliminary evaluation result, wherein the preliminary evaluation result comprises market risks, technical risks and policy risks and calculation weights corresponding to the three risks; And fifthly, setting a risk calculation interval by the summarizing module according to the preliminary evaluation result, and calculating the total startup risk according to the risk calculation interval.
2. The method for evaluating startup risk based on industry data analysis according to claim 1, wherein the preliminary domain module in the first step uses a text classification algorithm BERT to realize industry domain identification, establishes a plurality of domain dictionaries, and identifies a specific startup domain in the startup project by verifying related descriptive contents of the startup project through the domain dictionaries.
3. The method for evaluating startup risk based on industry data analysis according to claim 2, wherein in the second step, public data of national statistical bureau, industry association and professional database are adopted as industry data in different fields, the data supply module adopts ETL flow to clean, convert and load the original data, constructs a standardized startup risk field library, sets time for periodic update, and updates the content of the startup risk field library by adopting a common data update mode.
4. The method for entrepreneur risk assessment based on industry data analysis according to claim 3, wherein in the third step, the data supply module receives the specific entrepreneur domain output by the preliminary domain module, selects the entrepreneur data in the database according to the specific entrepreneur domain, classifies the entrepreneur data by using the existing large language model LLM, and obtains classification results of market data, technical data and policy data.
5. The method for evaluating startup risk based on industry data analysis according to claim 4, wherein the subdivision evaluation module in the fourth step comprises three sub-modules, each sub-module adopts different evaluation modes, and market risk, technical risk and policy risk are calculated by using market data, technical data and policy data respectively.
6. The startup risk assessment method based on industry data analysis according to claim 5, wherein the subdivision assessment module is specifically implemented as follows: market risk, namely adopting the market saturation in the market data as market risk calculation; ; ; Technical risk the technical risk is comprehensively evaluated by three aspects of technical maturity, research and development investment and patent number in technical data: Policy risk using the frequency of policy changes in the policy data as the policy risk, the calculation formula is as follows: ; the time interval in the above formula is defined as a person, the number of policy changes is the number of policy changes in the time interval, and in the above formula, 、 And To calculate the weights, the duty cycles of the different risks are defined.
7. The method for evaluating startup risk based on industry data analysis according to claim 6, wherein a specific implementation manner of calculating the weight is as follows, wherein the data volume of the startup risk field library is adopted as the calculation weight, and the calculation formula is as follows: ; ; the data volume ratio in the startup risk field library is used as the dividing basis of the calculation weight, so that the data of the startup risk field library can be more attached.
8. The method for evaluating startup risk based on industry data analysis according to claim 6, wherein in the fifth step, the summarization module obtains total startup risk according to the preliminary evaluation result provided by the subdivision evaluation module, and according to the set risk calculation interval, the calculation formula is as follows: ; in the above, the total startup risk is in the interval Is of low risk, in the interval In the interval for the risk of apoplexy Is a high risk.

Description

Entrepreneur risk assessment method based on industry data analysis Technical Field The invention relates to the technical field of startup risk assessment calculation, in particular to a startup risk assessment method based on industry data analysis. Background The traditional startup risk assessment method forms a certain system in long-term practice, but is faced with serious challenges brought by the digital age in terms of data dimension, analysis depth and dynamic adaptability. In the existing entrepreneur process, because of the problems of inaccurate industry data and unclear data sources, the existing entrepreneur cannot deal with different problems in the entrepreneur process, such as incapability of quickly determining the number of potential clients in the market, incapability of reasonably predicting technical risks according to the innovation capability of the quantitative entrepreneur process, incapability of quickly adapting to policy changes and the like. Disclosure of Invention The invention aims to provide a startup risk assessment method based on industry data analysis, which is characterized in that related industry data is acquired by means of a multi-channel data acquisition mechanism, a high-precision and traceable industry database is constructed, the accuracy of the data is guaranteed, the quick positioning and effectiveness verification of data sources can be realized, a double-layer risk assessment module and a data summarization module are built, and core assessment information required by startup planning is quickly output by combining a mature big data analysis model. On the basis, the matched target data are searched based on the database, the quantitative calculation of the startup risk is completed by adopting a statistical analysis method, and the accurate startup risk assessment based on the industry data is finally realized. In order to achieve the above purpose, the invention provides a startup risk assessment method based on industry data analysis, which comprises the following steps: setting a double-layer evaluation module and a summarization module, wherein the double-layer evaluation module comprises a preliminary field module and a subdivision evaluation module; acquiring industry data in different fields to form an entrepreneur risk field library, wherein a data supply module is arranged in the entrepreneur risk field library, and the data supply module normally updates the content of the entrepreneur risk field library according to the set updating time; step three, a user inputs entrepreneur related description content, a preliminary field module is combined with a data supply module, a specific field corresponding to the entrepreneur related description content in an entrepreneur risk field library is confirmed based on the entrepreneur related description content, the entrepreneur risk field library correspondingly outputs entrepreneur data of the specific field, and the data supply module classifies the acquired entrepreneur data; Inputting the classified startup data into a subdivision evaluation module to obtain a preliminary evaluation result, wherein the preliminary evaluation result comprises market risks, technical risks and policy risks and calculation weights corresponding to the three risks; And fifthly, setting a risk calculation interval by the summarizing module according to the preliminary evaluation result, and calculating the total startup risk according to the risk calculation interval. Preferably, the preliminary field module in the first step uses a text classification algorithm BERT to realize industry field recognition, establishes a plurality of field dictionaries, and recognizes a specific startup field in the startup project by checking related descriptive contents of the startup project through the field dictionaries. Preferably, in the second step, public data of a national statistical office, an industry association and a professional database are adopted as industry data of different fields, the data supply module adopts an ETL process to clean, convert and load the original data, a standardized startup risk field library is constructed, the time of regular update is set, and the content of the startup risk field library is updated by adopting a common data update mode. Preferably, in the third step, the data supply module receives the specific startup area output by the preliminary area module, selects startup data in the database according to the specific startup area, classifies the startup data by using the existing large language model LLM, and the classification result obtained is market data, technical data and policy data. Preferably, the subdivision evaluation module in the fourth step includes three sub-modules, each sub-module adopts different evaluation modes, and market risk, technical risk and policy risk are respectively calculated by using market data, technical data and policy data. Preferably, the subdivision eval