Electronic signature processing method and system for public accumulation fund business process data

Abstract

The invention relates to the technical field of data processing, and discloses an electronic signature processing method and system for public accumulation fund business process data, and the method comprises the steps: obtaining a target public accumulation fund contract in a PDF format; analyzing the target accumulation fund contract, and respectively determining the most profit data of each contract subject in the target accumulation fund contract and the compensation data of the target accumulation fund contract; extracting the most profit data and the compensation data in the target accumulation fund contract to form a fragmented contract, and calculating a standard hash value of the fragmented contract; repeatedly executing the following steps: performing first tampering on the most profitable data in the fragmented contract within a first preset range; performing second tampering on the compensation data in the fragmented contract in a second preset range; obtaining a plurality of tampered fragmented contracts based on the first tampering and the second tampering, and obtaining a tampering hash value of each tampered fragmented contract; and obtaining a tampering risk of the target accumulation fund contract based on a collision result of the standard hash value and any tampered hash value.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and more specifically, it relates to an electronic signature processing method and system for provident fund business process data. Background Art

[0002] With the exponential increase in computing power, the traditional electronic signature system based on the MD5 hash algorithm faces the risk of being brute-forced, that is, there is a possibility that provident fund business process files with different critical data generate the same MD5 hash value. Although existing electronic signature systems have been upgraded with new hash algorithms, for historical provident fund business process files based on the MD5 algorithm, there is still a potential risk of being tampered with or having been tampered with. Therefore, it is necessary to identify provident fund business process files with the risk of tampering and transfer them to manual for further processing. However, how to systematically identify the risk of tampering of provident fund business process files is a problem existing in the prior art. Summary of the Invention

[0003] The present invention provides an electronic signature processing method and system for provident fund business process data to solve the technical problems raised in the background art.

[0004] In a first aspect, the present invention provides an electronic signature processing method for provident fund business process data, including: Step 1, obtaining a target provident fund contract in PDF format; Step 2, parsing the target provident fund contract to respectively determine the most favorable data of each contract subject in the target provident fund contract and the compensation data of the target provident fund contract; Step 3, extracting the most favorable data and compensation data in the target provident fund contract to form a fragmented contract, and calculating the standard hash value of the fragmented contract; Step 4, repeatedly executing the following steps: Performing a first tampering on the most favorable data in the fragmented contract within a first preset range; Performing a second tampering on the compensation data in the fragmented contract within a second preset range; Obtaining a number of tampered fragmented contracts based on the first tampering and the second tampering, and obtaining the tampered hash value of each tampered fragmented contract; Step 5, obtaining the tampering risk of the target provident fund contract based on the collision result between the standard hash value and any one of the tampered hash values.

[0005] Further, determining the most favorable data of the contract subject includes: Obtain N data from the target provident fund contract and obtain the dependency relationships of the N data; where the dependency relationship means that if the i-th data is modified, the j-th data needs to be modified correspondingly to maintain the correct logical relationship of the data in the target provident fund contract, then there is a dependency relationship between the i-th data and the j-th data; Tamper with each character of the i-th data, and based on the dependency relationship, tamper with the data that has a dependency relationship with the i-th data correspondingly to obtain a tampered contract; Calculate the gain amount of each contract subject in the target provident fund contract based on the tampered contract; where the gain amount means, for any contract subject, the difference between the payment amount of the tampered contract and the payment amount of the target provident fund contract; Calculate the most profitable number of each contract subject for each data respectively, and the calculation formula is as follows: ; Where, represents the most profitable number of the i-th data for the s-th contract subject, represents the total number of tampered characters obtained by tampering with each character of the i-th data and based on the dependency relationship, represents the number of data that has a dependency relationship with the i-th data, represents the gain amount of the s-th contract subject; Based on the most profitable numbers of the s-th contract subject and the N data, select the data with the largest most profitable number for the s-th contract subject as the most profitable data of the s-th contract subject to obtain the most profitable data of each contract subject.

[0006] Furthermore, determine the compensation data of the target provident fund contract; where the compensation data represents the visual signature and visual seal in the target provident fund contract.

[0007] Furthermore, form a sharded contract, including the first extraction, the second extraction, and data fusion: The first extraction includes: extracting the most profitable data of the target provident fund contract, and the extraction includes: the visual area of the most profitable data in the target provident fund contract to obtain the first extraction data; The second extraction includes: extracting the compensation data of the target provident fund contract, and the extraction includes: determining the allowed writing position and allowed stamping position of the visual signature and visual seal in the target provident fund contract based on manual experience; extracting the allowed writing position in the visual area of the target provident fund contract; extracting the allowed stamping position in the visual area of the target provident fund contract to obtain the second extraction data; Data fusion includes: initializing to generate a blank page in PDF format, where the blank page is the same size as the target provident fund contract, and fitting the first extracted data and the second extracted data to the blank page to obtain a fragmented contract.

[0008] Furthermore, the first preset range includes: obtaining the maximum and minimum values of each most favorable data in the target provident fund contract within a historical time period, and using the maximum and minimum values as the maximum and minimum values of the first preset range corresponding to the most favorable data.

[0009] Furthermore, the second preset range includes: Based on the visual signature and visual seal, establishing the constraint conditions for the second preset range, including: The first constraint: the rewritten area of the visual signature and visual seal is greater than or equal to the preset area threshold; The second constraint: the position of the visual signature does not exceed the allowed writing position; The third constraint: the position of the visual seal does not exceed the allowed stamping position; Based on the combined constraints of the first constraint, the second constraint, and the third constraint, the second preset range is obtained.

[0010] Furthermore, based on the first tampering and the second tampering, a number of tampered fragmented contracts are obtained, including performing the following steps for each contract subject: Performing the first tampering: obtaining the data precision of the most favorable data of the s-th contract subject, and performing gradient descent based on the maximum value of the s-th contract subject within the first preset range in the smallest unit of the data precision until reaching the minimum value of the s-th contract subject within the first preset range or until the standard hash value and the tampered hash value collide successfully; Performing the second tampering: traversing the visual signature and visual seal within the second preset range with pixel-level precision to obtain K tampering compensation schemes; In the gradient order, for each gradient of the first tampering, match the K tampering compensation schemes.

[0011] Furthermore, both the standard hash value and the tampered hash value are calculated based on the MD5 hash algorithm.

[0012] Furthermore, the tampering risk of the target provident fund contract is calculated based on the following formula: ; Wherein, represents the tampering risk of the target provident fund contract, represents the number of contract subjects of the target provident fund contract, represents the index of, represents the The value of the most favorable data corresponding to the collision of contract entities, indicating the value of the most favorable data corresponding to the nth contract entity in the target provident fund contract,

[0013] indicating the operation of taking the absolute value. A data acquisition module for obtaining a target provident fund contract in PDF format; A data parsing module for parsing the target provident fund contract to respectively determine the most favorable data of each contract entity in the target provident fund contract and the compensation data of the target provident fund contract; A contract sharding module for extracting the most favorable data and compensation data in the target provident fund contract to form a sharded contract and calculating the standard hash value of the sharded contract; A hash collision module for repeatedly executing the following steps: Performing a first tampering on the most favorable data in the sharded contract within a first preset range; Performing a second tampering on the compensation data in the sharded contract within a second preset range; Obtaining a number of tampered sharded contracts based on the first tampering and the second tampering and obtaining the tampered hash value of each tampered sharded contract; A risk quantification module for obtaining the tampering risk of the target provident fund contract based on the collision result between the standard hash value and any one of the tampered hash values.

[0014] The beneficial effects of the present invention are as follows: 1. By performing controlled tampering on the most favorable data and compensation data in the target provident fund contract and establishing an MD5 hash collision detection mechanism, the present invention effectively identifies and quantifies the potential tampering risks of the contract by adopting a gradient and refined tampering method for the most favorable data of the contract entity and traversing multiple schemes at the pixel level for the visual signature and seal.

[0015] 2. Without changing the existing system architecture, the present invention realizes the early warning and classification processing of the risks of historical legacy files, enhancing the security management and risk prevention and control capabilities of the entire business process. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 is a flowchart of an electronic signature processing method for provident fund business process data according to the present invention; Figure 2 is a module diagram of an electronic signature processing system for provident fund business process data according to the present invention. DETAILED DESCRIPTION

[0017] Reference will now be made to exemplary embodiments to discuss the subject matter described herein. It should be understood that discussing these embodiments is only to enable those skilled in the art to better understand and thus implement the subject matter described herein. Without departing from the scope of protection of the content of this specification, changes can be made to the functions and arrangements of the elements discussed. Each example may omit, substitute, or add various processes or components as needed. Additionally, the features described relative to some examples can also be combined in other examples.

[0018] As Figures 1 to 2 shown, an electronic signature processing method for provident fund business process data includes: Step 1, obtain a target provident fund contract in PDF format; Step 2, parse the target provident fund contract, and respectively determine the most favorable data of each contract subject in the target provident fund contract and the compensation data of the target provident fund contract; Step 3, extract the most favorable data and compensation data in the target provident fund contract to form a sharded contract, and calculate the standard hash value of the sharded contract; Step 4, repeatedly execute the following steps: Perform a first tampering on the most favorable data in the sharded contract within a first preset range; Perform a second tampering on the compensation data in the sharded contract within a second preset range; Based on the first tampering and the second tampering, obtain a number of tampered sharded contracts, and obtain the tampered hash value of each tampered sharded contract; Step 5, based on the collision result between the standard hash value and any one of the tampered hash values, obtain the tampering risk of the target provident fund contract.

[0019] It should be noted that in a typical provident fund contract, its content includes text, signature, and seal. Among them, the text content is displayed in a fixed font, font size, and predetermined position, while although the signature and seal content also use a fixed font and font size, their display positions are random within the permitted area. Therefore, when modifying the text content causes the hash value to change, the position of the signature and seal content can be offset within the permitted area to randomly compensate for the hash value change caused by the text modification, thereby achieving hash collision.

[0020] In an embodiment of the present invention, determining the most favorable data of a contract subject includes: Obtain N data in the target provident fund contract, and obtain the dependency relationship of the N data; where the dependency relationship means that if the i-th data is modified, then the j-th data needs to be correspondingly modified to keep the logical relationship of the data in the target provident fund contract correct, then there is a dependency relationship between the i-th data and the j-th data; Tamper with each character of the i-th data, and based on the dependency relationship, tamper with the data that has a dependency relationship with the i-th data accordingly to obtain a tampered contract; Calculate the gain amount of each contract subject in the target provident fund contract based on the tampered contract; wherein, the gain amount represents, for any contract subject, the difference between the payment amount of the tampered contract and the payment amount of the target provident fund contract; Calculate the most profitable number of each contract subject for each data respectively, and the calculation formula is as follows: ; Wherein, represents the most profitable number of the i-th data for the s-th contract subject, represents the total number of tampered characters obtained by tampering with each character of the i-th data and based on the dependency relationship, represents the number of data that has a dependency relationship with the i-th data, represents the -th gain amount of the contract subject; Based on the most profitable numbers of the s-th contract subject and N data, select the data with the largest most profitable number for the s-th contract subject as the most profitable data of the s-th contract subject, so as to obtain the most profitable data of each contract subject.

[0021] Specifically, to solve the problem that the computational complexity grows exponentially when facing numerous data items in the contract. Since this solution performs risk detection based on the brute-force collision method, each additional change in a data item will significantly increase the computational complexity. Therefore, it is necessary to efficiently screen the data modification process. In this solution, by extracting all data items in the contract and establishing the dependency relationship between them (that is, when modifying a certain data item, the related data items need to be adjusted synchronously to ensure the overall logic of the contract is correct), the operation of controlling the tampering of each character of each data item is realized. At the same time, calculate the "gain amount" (that is, the difference between the tampered contract and the original contract payment amount) and the cumulative number of modified characters generated during the tampering process for each item, so as to obtain the "most profitable number" of each item. This method actually simulates the psychology of the tamperer who strives to obtain the maximum modification benefit at the minimum modification cost. By screening out the data item with the largest "most profitable number" as the key data, the overall computational burden is effectively reduced, and at the same time, the efficiency and accuracy of detecting potential tampering risks are improved.

[0022] In an embodiment of the present invention, determine the compensation data of the target provident fund contract; wherein, the compensation data represents the visual signature and visual seal in the target provident fund contract.

[0023] Specifically, the compensation data refers to the visual signature and visual seal in the provident fund contract. In a typical provident fund contract, the text content is displayed in a fixed font, font size, and fixed position. Therefore, any minor modification to the text content will cause an obvious change in the hash value. Although the signature and seal also use a fixed font and font size, their specific positions within the permitted area are random. For this reason, if one wants to modify the most favorable data, by appropriately adjusting the positions of the signature and seal, the change in the hash value caused by the modification of the most favorable data can be randomly compensated, thereby achieving hash value collision detection.

[0024] In an embodiment of the present invention, a sharded contract is formed, including first extraction, second extraction, and data fusion: The first extraction includes: extracting the most favorable data of the target provident fund contract, and the extraction includes: the most favorable data in the visual area of the target provident fund contract to obtain the first extraction data; The second extraction includes: extracting the compensation data of the target provident fund contract, and the extraction includes: determining the permitted writing position and permitted stamping position of the visual signature and visual seal in the target provident fund contract based on manual experience; extracting the permitted writing position in the visual area of the target provident fund contract; extracting the permitted stamping position in the visual area of the target provident fund contract to obtain the second extraction data; Data fusion includes: initializing and generating a blank page in PDF format, where the blank page is the same size as the target provident fund contract, and fitting the first extraction data and the second extraction data to the blank page to obtain the sharded contract.

[0025] It should be noted that there is a combinatorial property in hash value calculation, that is, for a fixed and unchanging part of a document (denoted as document 0), if two document parts with the same position size but different contents (denoted as document 1 and document 2) have the same hash value, then after splicing document 0 with document 1 or document 2 at the same position, that is, forming 1+0 and 2+0, their hash values will also remain the same. Utilizing this property, by dividing the contract into an unchanging part and a controllable tampering part: during the brute-force collision detection process, the unchanging part is masked, and the hash value of the controllable tampering part is calculated, thereby reducing the amount of computation and improving the computation efficiency.

[0026] In an embodiment of the present invention, the first preset range includes: obtaining the maximum value and minimum value of each most favorable data in the target provident fund contract in the historical time period, and using the maximum value and minimum value as the maximum value and minimum value of the first preset range of the corresponding most favorable data.

[0027] Specifically, the first preset range is determined by obtaining the maximum and minimum values of each "most favorable data" in the contract within the historical time period. That is to say, for each key data item, we use its highest and lowest values recorded in the historical data as the upper and lower bounds of the acceptable modification of the data respectively, thereby limiting the range of variation allowed for the data during the controlled tampering process. By strictly limiting the modification range within a reasonable interval, the number of hash values that need to be calculated during the brute-force collision process can be significantly reduced, thus improving the overall operation efficiency. The first preset range not only provides a quantitative standard for subsequent gradient-based controlled tampering but also reduces the computational burden.

[0028] In an embodiment of the present invention, the second preset range includes: Establishing constraint conditions for the second preset range based on the visual signature and the visual seal, including: The first constraint: The rewritten area of the visual signature and the visual seal is greater than or equal to the preset area threshold; The second constraint: The position of the visual signature does not exceed the allowed writing position; The third constraint: The position of the visual seal does not exceed the allowed stamping position; The second preset range is obtained based on the combined constraints of the first constraint, the second constraint, and the third constraint.

[0029] It should be noted that the second preset range is mainly used to constrain the position adjustment of the visual signature and the visual seal to ensure that effective hash collision detection can be performed during the controlled tampering process without deviating from the actual application rules of the contract. Among them, in a typical provident fund contract, the seal is usually stamped above the signature. Therefore, when adjusting the positions of the signature and the seal, it is necessary to ensure that the overlapping area of the two reaches at least the threshold value to ensure that the adjusted contract still conforms to the business practice convention. In addition, the second constraint and the third constraint are respectively used to ensure that the position adjustment of the signature and the seal does not exceed the specified writing area and stamping area of the contract. The role of these constraints is to ensure that the adjusted contract still conforms to the regulations of the provident fund business process and avoid invalidation or loss of legal effect of the contract due to excessive adjustment.

[0030] In an embodiment of the present invention, based on the first tampering and the second tampering, a number of tampered shard contracts are obtained, including performing the following steps for each contract subject: Performing the first tampering: Obtaining the data precision of the most favorable data of the s-th contract subject, and performing gradient descent based on the maximum value of the s-th contract subject within the first preset range in the smallest unit of the data precision until reaching the minimum value of the s-th contract subject within the first preset range or until the standard hash value and the tampered hash value collide successfully; Perform the second tampering: Traverse the visual signature and visual seal at the pixel-level accuracy within the second preset range to obtain K tampering compensation schemes; For each gradient of the first tampering, match the K tampering compensation schemes in the order of gradient.

[0031] Specifically, two different tampering strategies are adopted for the text data and the compensation data (signature and seal), and all possible collision situations are traversed by matching their different adjustment combinations.

[0032] First tampering: First, obtain the most favorable data of each contract subject in the target contract and determine its data accuracy (i.e., the minimum adjustable unit. For example, the minimum unit of the most favorable data 20.12 is 0.01). Then, according to the first preset range, modify the most favorable data in a gradient descent manner, that is, gradually decrease from the maximum allowable value to the minimum allowable value, or until a collision is successful. This method simulates the strategy of the tamperer to pursue the smallest change but still obtain the maximum benefit in actual operation, and at the same time reduces the amount of calculation. While adjusting the text data, within the second preset range, adjust the positions of the signature and the seal to make up for the change in the hash value caused by the text tampering. By adopting a pixel-level traversal method (without changing the shape and size of the visual signature and seal, only changing the position), that is, within the allowable writing and stamping areas, traverse the possible tampering compensation schemes. After completing the first tampering (text modification) and the second tampering (signature and seal adjustment), the present invention matches each text modification scheme with the K different compensation schemes in the order of gradient, and calculates the hash value after tampering.

[0033] The present invention does not rely on a single data adjustment to achieve a hash collision, but combines the precise gradient modification of the text data with the flexible adjustment of the visual data (signature and seal), so that even if the text part is tampered with, the change in the hash value can be compensated by the fine adjustment of the signature and the seal, thereby more effectively detecting possible malicious tampering. In addition, under the limitation of limited computing resources, this method greatly reduces the unnecessary amount of hash calculation and improves the efficiency of tampering risk detection.

[0034] In an embodiment of the present invention, both the standard hash value and the tampered hash value are calculated based on the MD5 hash algorithm.

[0035] In an embodiment of the present invention, the tampering risk of the target provident fund contract is calculated based on the following formula: ; Wherein, represents the tampering risk of the target provident fund contract, represents the number of contract subjects of the target provident fund contract, represents the index of Indicates the value of the most favorable data corresponding to the successful collision of the th contract entity. Indicates the value of the most favorable data corresponding to the th contract entity in the target provident fund contract. Indicates the operation of taking the absolute value.

[0036] In one embodiment of the present invention, the gain situation of the contract entity is calculated by a formula to quantify the risk degree of the tampering behavior. At the same time, if the tampering risk of the target provident fund contract is greater than or equal to a preset threshold, the target provident fund contract will be transferred to manual processing.

[0037] An electronic signature processing system for provident fund business process data, applied to any of the above methods, includes: A data acquisition module for acquiring a target provident fund contract in PDF format; A data parsing module for parsing the target provident fund contract to respectively determine the most favorable data of each contract entity in the target provident fund contract and the compensation data of the target provident fund contract; A contract sharding module for extracting the most favorable data and compensation data in the target provident fund contract to form a sharded contract and calculating the standard hash value of the sharded contract; A hash collision module for repeatedly executing the following steps: Performing a first tampering on the most favorable data in the sharded contract within a first preset range; Performing a second tampering on the compensation data in the sharded contract within a second preset range; Obtaining a number of tampered sharded contracts based on the first tampering and the second tampering, and acquiring the tampered hash value of each tampered sharded contract; A risk quantification module for obtaining the tampering risk of the target provident fund contract based on the collision result between the standard hash value and any one of the tampered hash values.

[0038] The innovations of the present invention are as follows: 1. A large number of historical provident fund contracts still use MD5 for signature verification. The present invention can seamlessly adapt to these existing files without changing the original electronic signature archiving mechanism.

[0039] 2. The easy collision property of MD5 enables the present invention to perform controllable modification on contract data through constructive tampering and detect possible tampering risks.

[0040] The above has described the embodiments of this embodiment, but this embodiment is not limited to the above specific implementation manners. The above specific implementation manners are merely illustrative rather than restrictive. Under the inspiration of this embodiment, those of ordinary skill in the art can also make many forms, all of which fall within the protection scope of this embodiment.

Claims

1. An electronic signature processing method for provident fund business process data, characterized in that, Including: Step 1: Obtain the target provident fund contract in PDF format; Step 2: Analyze the target provident fund contract to respectively determine the most favorable data of each contract subject in the target provident fund contract and the compensation data of the target provident fund contract; Step 3: Extract the most favorable data and compensation data in the target provident fund contract to form a fragmented contract, and calculate the standard hash value of the fragmented contract; Step 4: Repeat the following steps: Perform the first tampering on the most favorable data in the fragmented contract within the first preset range; Perform the second tampering on the compensation data in the fragmented contract within the second preset range; Obtain a number of tampered fragmented contracts based on the first tampering and the second tampering, and obtain the tampered hash value of each tampered fragmented contract; Step 5: Based on the collision result between the standard hash value and any one of the tampered hash values, obtain the tampering risk of the target provident fund contract.

2. The electronic signature processing method for provident fund business process data according to claim 1, characterized in that Determine the most favorable data of the contract subject, including: Obtain N data in the target provident fund contract and obtain the dependency relationship of the N data; among them, the dependency relationship means that if the i-th data is modified, the j-th data needs to be modified correspondingly to keep the logical relationship of the data in the target provident fund contract correct, then there is a dependency relationship between the i-th data and the j-th data; Tamper with each character of the i-th data, and based on the dependency relationship, tamper with the data that has a dependency relationship with the i-th data correspondingly to obtain a tampered contract; Calculate the gain amount of each contract subject in the target provident fund contract based on the tampered contract; among them, the gain amount means, for any contract subject, the difference between the payment amount of the tampered contract and the payment amount of the target provident fund contract; Calculate the most favorable number of each contract subject for each data respectively, and the calculation formula is as follows: ； Among them, represents the most favorable number of the i-th data for the s-th contract entity, represents the total number of tampered characters obtained by tampering with each character of the i-th data and based on the dependency relationship, represents the number of data that has a dependency relationship with the i-th data, represents the gain amount of the s-th contract entity; Based on the most favorable number of the s-th contract subject and the N data, select the data with the largest most favorable number of the s-th contract subject as the most favorable data of the s-th contract subject, so as to obtain the most favorable data of each contract subject.

3. The electronic signature processing method for provident fund business process data according to claim 2, characterized in that, Determine the compensation data of the target provident fund contract; among them, the compensation data represents the visual signature and visual seal in the target provident fund contract.

4. An electronic signature processing method for provident fund business process data according to claim 3, characterized in that Form a fragmented contract, including the first extraction, the second extraction and data fusion: The first extraction includes: extracting the most favorable data of the target provident fund contract, and the extraction includes: the most favorable data in the visual area of the target provident fund contract to obtain the first extraction data; The second extraction includes: extracting the compensation data of the target provident fund contract, and the extraction includes: determining the allowed writing position and allowed stamping position of the visual signature and visual seal in the target provident fund contract based on manual experience; extracting the allowed writing position in the visual area of the target provident fund contract; extracting the allowed stamping position in the visual area of the target provident fund contract to obtain the second extraction data; Data fusion includes: initializing and generating a blank page in PDF format, the blank page has the same size as the target provident fund contract, and fitting the first extraction data and the second extraction data to the blank page to obtain a fragmented contract.

5. An electronic signature processing method for provident fund business process data according to claim 4, characterized in that, The first preset range includes: obtaining the maximum and minimum values of each most favorable data in the target provident fund contract within the historical time period, and using the maximum and minimum values as the maximum and minimum values of the first preset range of the corresponding most favorable data.

6. The electronic signature processing method for provident fund business process data according to claim 5, characterized in that, The second preset range includes: Establishing the constraint conditions for the second preset range based on the visual signature and visual seal, including: The first constraint: the rewritten area of the visual signature and the visual seal is greater than or equal to the preset area threshold; The second constraint: the position of the visual signature does not exceed the allowed writing position; The third constraint: the position of the visual seal does not exceed the allowed stamping position; The second preset range is obtained based on the combined constraints of the first constraint, the second constraint, and the third constraint.

7. An electronic signature processing method for provident fund business process data according to claim 6, characterized in that, Based on the first tampering and the second tampering, a number of tampered shard contracts are obtained, including performing the following steps for each contract subject: Performing the first tampering: obtaining the data precision of the most favorable data of the s-th contract subject, and performing gradient descent based on the maximum value of the s-th contract subject within the first preset range in the minimum unit of the data precision until reaching the minimum value of the s-th contract subject within the first preset range or until the standard hash value and the tampered hash value collide successfully; Performing the second tampering: traversing the visual signature and the visual seal within the second preset range with pixel-level precision to obtain K tampering compensation schemes; For each gradient of the first tampering, match the K tampering compensation schemes in the gradient order.

8. An electronic signature processing method for provident fund business process data according to claim 7, characterized in that Both the standard hash value and the tampered hash value are calculated based on the MD5 hash algorithm.

9. An electronic signature processing method for provident fund business process data according to claim 8, characterized in that, The tampering risk of the target provident fund contract is calculated based on the following formula: ； Among them, represents the tampering risk of the target provident fund contract, represents the number of contract parties of the target provident fund contract, represents the index of represents the value of the most favorable data corresponding when the th contract party collision is successful, represents the value of the most favorable data corresponding to the th contract party in the target provident fund contract, represents the operation of taking the absolute value.

10. An electronic signature processing system for provident fund business process data, characterized in that, Applied to the method described in any one of claims 1-9, it includes: A data acquisition module for obtaining the target provident fund contract in PDF format; A data parsing module for parsing the target provident fund contract to respectively determine the most favorable data of each contract subject in the target provident fund contract and the compensation data of the target provident fund contract; A contract sharding module for extracting the most favorable data and the compensation data in the target provident fund contract to form a shard contract and calculating the standard hash value of the shard contract; A hash collision module for repeatedly performing the following steps: Performing the first tampering on the most favorable data in the shard contract within the first preset range; Performing the second tampering on the compensation data in the shard contract within the second preset range; Obtaining a number of tampered shard contracts based on the first tampering and the second tampering, and obtaining the tampered hash value of each tampered shard contract; A risk quantification module for obtaining the tampering risk of the target provident fund contract based on the collision result of the standard hash value and any one of the tampered hash values.

Images