CN-121436995-B - False trade monitoring method and system based on data cross verification

CN121436995BCN 121436995 BCN121436995 BCN 121436995BCN-121436995-B

Abstract

The invention provides a false trade monitoring method and system based on data cross verification, which belong to the technical field of information monitoring and comprise a data acquisition module, a dynamic blockchain storage module, an intelligent cross verification module, a self-learning Xi Geyao module, a real-time early warning processing module and a real-time early warning processing module, wherein the data acquisition module is used for acquiring multi-source trade data containing enterprise unique codes and dynamic characteristic identifiers, the dynamic blockchain storage module is used for storing offline trade data in a alliance chain architecture and ensuring that the offline trade data cannot be tampered, the intelligent cross verification module is used for cross comparison of the data through an industry characteristic anomaly algorithm, the self-learning Xi Geyao module is used for verifying compliance based on an initial rule and a dynamic model, and the real-time early warning processing module is used for generating early warning and triggering business freezing and improving false trade identification capability. The invention solves the problems of data island, asymmetric information and the like in the traditional trade, remarkably improves the prevention capability of false trade and lays a technical foundation for trade safety.

Inventors

ZHAO YUANBO
WANG XIAOBO
WANG CHAO
SONG JINWEN
YU SHOUSHUI
WANG JINGYI
LIU HAITAO

Assignees

中国华电集团产融控股有限公司
华电商业保理(天津)有限公司

Dates

Publication Date: 20260512
Application Date: 20251104

Claims (8)

1. The false trade monitoring system based on the data cross verification is characterized by comprising a data acquisition module, a dynamic blockchain storage module, an intelligent cross verification module, a self-learning contract module and a real-time early warning processing module, wherein the modules realize cooperative communication through an encrypted data structure; The data acquisition module is used for acquiring real-time multi-source trade data, wherein the multi-source trade data comprises enterprise unique codes and dynamic characteristic identifiers, and covers structured business data autonomously provided by enterprises, unstructured public data issued by a supervision mechanism and semi-structured associated data of a third party authority; The dynamic blockchain storage module adopts a alliance chain architecture to deploy an endorsement node, an accounting node and a cross-chain gateway, is used for storing offline trade data updated according to a dynamic period, ensures that the data cannot be tampered through hash encryption, a time stamp and a node signature, and realizes data synchronization with an industry vertical chain through the cross-chain gateway; The intelligent cross verification module is used for calling the offline trade data of the dynamic blockchain storage module, carrying out dimension association comparison with the real-time multisource trade data, calculating comprehensive anomaly degree through an anomaly degree quantization algorithm fusing industry characteristics, and identifying data inconsistency and logic conflict; the intelligent cross verification module comprises: The characteristic association unit takes the enterprise unique code as a core and establishes a dynamic association map of trade data and financing data and warehousing data; The abnormal calculation unit is used for fusing the abnormal degree quantization algorithm formula of the industry characteristics, wherein the abnormal degree quantization algorithm formula is as follows: ; Wherein, the 、、 Are all weight coefficients dynamically adjusted based on industry characteristics The method comprises the steps of (1) setting field matching rate X omega 2 of the industry, wherein omega 1 、ω 2 is an industry characteristic coefficient, T is data timeliness, T=duration exceeding industry allowable deviation/industry maximum reasonable deviation duration, T is less than or equal to 1;R is data relevance, R is trade financing data matching rate X lambda 1 + trade warehouse data matching rate X lambda 2 , lambda 1 、λ 2 is a dynamic relevance coefficient, and judging to be suspected false trade when D reaches an industry preset threshold; The self-learning contract module is internally provided with an initial compliance rule base, a dynamic rule updating model is generated through historical verification result training, compliance of real-time data and offline data is automatically verified, and a compliance verification result with confidence is output; the real-time early warning disposal module is used for generating early warning grades by combining the comprehensive abnormal degree and the compliance verification result, triggering a service freezing interface of the corresponding grade, and writing early warning and disposal records into the dynamic block chain storage module to form a closed loop.
2. The data cross-validation based dummy trade monitoring system of claim 1, wherein said data acquisition module comprises: The enterprise terminal acquisition unit is used for configuring the self-adaptive interface to adapt to different enterprise ERP systems and acquiring the structured business data containing dynamic characteristic representation in real time; The supervision terminal acquisition unit extracts information in unstructured public data through OCR recognition and semantic analysis technology; and the third party acquisition unit adopts an API gateway to aggregate the semi-structured association data, performs format conversion and feature extraction, and then synchronizes the semi-structured association data to the monitoring system.
3. The data cross-validation based dummy trade monitoring system of claim 1, wherein in said dynamic blockchain storage module federated chain architecture, comprising: the endorsement node is deployed by a trade participant and a supervision mechanism respectively, and adopts a role-based authority control mechanism; the accounting node adopts a distributed storage cluster and divides storage partitions according to data sensitivity level; the cross-chain gateway realizes data interaction and verification with energy and financial vertical industry chains through an atomic exchange protocol.
4. The data cross-validation based false trade monitoring system of claim 1, wherein said self-learning contract module comprises: the rule base unit is used for storing trade flow rules, production and fusion association rules and industry rules; The training unit is used for training the historical verification data by adopting a gradient descent algorithm to generate a rule adjustment coefficient; And the execution unit is used for verifying the data compliance based on the initial rule and the adjustment coefficient and outputting a verification result containing 0-1 interval confidence coefficient.
5. A method for monitoring false trade based on data cross verification, comprising the steps of: s1, collecting real-time multi-source trade data, wherein the multi-source trade data comprises enterprise unique codes and dynamic characteristic identifiers, and covers structured business data, unstructured public data and semi-structured associated data; S2, storing offline trade data through a dynamic blockchain storage module, wherein the offline trade data is updated according to a dynamic period, hash encryption, a time stamp and a node signature are adopted to ensure that the offline trade data cannot be tampered, and the offline trade data are synchronized with energy and financial vertical industry chains through a cross-chain gateway; S3, constructing dynamic association graphs of trade data, financing data and storage data by taking enterprise unique codes as cores, and cleaning and extracting features of offline trade data; S4, calling offline trade data in the dynamic association map to carry out cross comparison with the real-time multisource trade data, and calculating comprehensive anomaly degree D through an anomaly degree quantization algorithm fusing industry characteristics; In S4, the algorithm for quantifying the degree of abnormality of the fusion industry feature includes: S41, calculating data consistency C, wherein C=basic field matching rate x omega 1 +industry set field matching rate x omega 2 , wherein omega 1 、ω 2 is an industry characteristic coefficient; s42, calculating the timeliness T of the data, wherein T=the time length exceeding the allowable deviation of the industry/the maximum reasonable deviation time length of the industry, and T is less than or equal to 1; S43, calculating data relevance R, wherein R=trade financing data matching rate x lambda 1 +trade warehouse data matching rate x lambda 2 , and lambda 1 、λ 2 is a dynamic relevance coefficient; S44, dynamically distributing alpha, beta and gamma weights according to industry types, substituting the weights into a formula to calculate ; S5, automatically verifying compliance of the real-time data and the offline data based on an initial compliance rule base and a dynamic rule updating model of the self-learning contract module, and outputting a compliance verification result with confidence; and S6, generating an early warning grade by combining the comprehensive anomaly degree D and the compliance verification result, triggering a corresponding service freezing interface, and writing an early warning and treatment record into a dynamic blockchain storage module.
6. A method of monitoring a fraudulent trade based on data cross-validation according to claim 5, wherein in S1, said collecting real-time multisource trade data includes: s11, collecting structured business data from an ERP system of an enterprise through a self-adaptive interface, and adding a dynamic characteristic identifier; S12, extracting unstructured data information from the public file of the supervision institution through OCR recognition and semantic analysis technology; S13, aggregating the semi-structured association data of the third party through the API gateway, converting the format, extracting the characteristics and synchronizing the characteristics to the monitoring system.
7. The method for monitoring false trade based on data cross-validation according to claim 5, wherein in S2, specifically comprising: s21, the endorsement node performs signature verification on the offline trade data, and generates a hash value and a time stamp after the offline trade data passes the signature verification; s22, the accounting node is stored to the corresponding partition according to the data sensitivity level, and the cross-chain gateway is synchronously verified with the energy industry chain and the financial industry chain through an atomic exchange protocol; S23, dynamically adjusting the storage period according to the data updating frequency, and shortening the updating interval by the high-frequency fluctuation data.
8. The method of claim 5, wherein in S6, the pre-alarm level comprises: S61, if D is more than or equal to an industry high threshold value and the compliance verification confidence is more than or equal to 0.8, generating a first-level early warning and triggering a core service freezing interface; s62, if the threshold value in the industry is less than or equal to D < the high threshold value in the industry or the compliance verification confidence coefficient is 0.5-0.8, generating a second-level early warning and triggering an associated service freezing interface; s63, if the industry low threshold value is less than or equal to D < industry medium threshold value or compliance verification confidence coefficient is 0.2-0.5, generating three-level early warning and pushing to a management end; And S64, finally, writing the early warning information, the treatment record and the verification basis into the dynamic blockchain storage module through the node signature.

Description

False trade monitoring method and system based on data cross verification Technical Field The invention relates to the technical field of information monitoring, in particular to a false trade monitoring method and system based on data cross verification. Background The trade is a core component of economic activity and covers various behaviors such as commodity exchange, buying and selling transaction and the like in domestic and international trade, and the trade is smoothly carried out to effectively transfer and verify multi-dimensional data such as contracts, logistics, financing, supervision and the like generated by trade participators, supervision authorities and third party authorities, and the authenticity and relevance of the data are directly related to trade safety. In current traditional trade monitoring, multisource trade data scatter storage forms "data island" in the independent system of different main parts, and cross main part information sharing difficulty leads to there is serious information asymmetry between the trade participators, and false trade action (like fake voucher, false financing) is difficult to be in time discerned, and prior art is weak to false trade's precaution ability, can't satisfy trade safety guarantee demand. Therefore, there is a need for a method and system for monitoring a dummy trade that can improve the accuracy and effectiveness of the monitoring of the dummy trade. Disclosure of Invention The invention aims to overcome the defects of the prior art, and provides a false trade monitoring method and system based on data cross verification, which solve the problems of data island, information asymmetry and the like in the traditional trade, remarkably improve the prevention capability of false trade and lay a technical foundation for trade safety. In order to achieve the above object, the present invention provides the following solutions: The invention aims at providing a false trade monitoring system based on data cross verification, which comprises a data acquisition module, a dynamic blockchain storage module, an intelligent cross verification module, a self-learning contract module and a real-time early warning processing module, wherein the modules realize cooperative communication through an encrypted data structure; The data acquisition module is used for acquiring real-time multi-source trade data, wherein the multi-source trade data comprises enterprise unique codes and dynamic characteristic identifiers, and covers structured business data autonomously provided by enterprises, unstructured public data issued by a supervision mechanism and semi-structured associated data of a third party authority; The dynamic blockchain storage module adopts a alliance chain architecture to deploy an endorsement node, an accounting node and a cross-chain gateway, is used for storing offline trade data updated according to a dynamic period, ensures that the data cannot be tampered through hash encryption, a time stamp and a node signature, and realizes data synchronization with an industry vertical chain through the cross-chain gateway; The intelligent cross verification module is used for calling the offline trade data of the dynamic blockchain storage module, carrying out dimension association comparison with the real-time multisource trade data, calculating comprehensive anomaly degree through an anomaly degree quantization algorithm fusing industry characteristics, and identifying data inconsistency and logic conflict; The self-learning contract module is internally provided with an initial compliance rule base, a dynamic rule updating model is generated through historical verification result training, compliance of real-time data and offline data is automatically verified, and a compliance verification result with confidence is output; the real-time early warning disposal module is used for generating early warning grades by combining the comprehensive abnormal degree and the compliance verification result, triggering a service freezing interface of the corresponding grade, and writing early warning and disposal records into the dynamic block chain storage module to form a closed loop. Preferably, the data acquisition module comprises: The enterprise terminal acquisition unit is used for configuring the self-adaptive interface to adapt to different enterprise ERP systems and acquiring the structured business data containing dynamic characteristic representation in real time; The supervision terminal acquisition unit extracts information in unstructured public data through OCR recognition and semantic analysis technology; and the third party acquisition unit adopts an API gateway to aggregate the semi-structured association data, performs format conversion and feature extraction, and then synchronizes the semi-structured association data to the monitoring system. Preferably, in the federated chain architecture of the dynamic blockchain storage module, the method