Listed company suspected financial counterfeiting behavior insight discrimination system

Abstract

The invention discloses a listed company suspected financial counterfeiting behavior insight discrimination system, and the system comprises: an information crawling module which is used for obtainingthe financial information and non-financial information of a listed company from the Internet; a condition verification module which is used for verifying financial information and non-financial information of the listed company according to a preset first early warning condition and a preset second early warning condition; a scoring module which is used for scoring financial conditions of listedcompanies; a first judgment module which is used for comparing the financial condition score of the listed company with a preset first threshold value to obtain a financial counterfeiting judgment result of the listed company; and a second judgment module which is used for continuously adopting the logistic regression model to calculate the financial counterfeiting probability of the listed company when the first judgment module obtains that the financial counterfeiting judgment result of the listed company is false, so as to obtain the financial counterfeiting judgment result of the listed company. According to the method, the logic rules of the accountant are combined with the machine learning model, so that financial counterfeiting behaviors can be identified more accurately.

Description

technical field [0001] The invention relates to the technical field of data processing. More specifically, the present invention relates to a system for insight and discrimination of suspected financial fraud behaviors of listed companies. Background technique [0002] At present, due to various factors such as audit scope limitations and audit time constraints, it is sometimes difficult for auditors to detect financial fraud in listed companies; and it is often difficult to analyze whether a company has financial fraud in the market. The information is considered together. The existing financial fraud analysis technologies and models on the market are often built on the basis of machine learning, but it is difficult to explain the effectiveness and rationality of the model from an accounting perspective. Therefore, how to combine the logical rules of accounting with machine learning models is the focus of current research in the field of financial fraud in listed companie...

AI Technical Summary

Problems solved by technology

[0002] At present, due to various factors such as audit scope limitations and audit time constraints, it is sometimes difficult for auditors to detect financial fraud in listed companies; and it is often difficult to analyze whether a company has financial fraud in the market. information combined

The existing financial fraud analysis technologies and models on the market are often built on the basis of machine learning, but it is difficult to explain the effectiveness and rationality of the model from an accounting perspective

Images