CN114880331A

CN114880331A - Financial data accurate tracing method based on block chain MPT tree

Info

Publication number: CN114880331A
Application number: CN202210694378.0A
Authority: CN
Inventors: 李保珍; 韩占校; 张亭亭; 王雷
Original assignee: NANJING AUDIT UNIVERSITY
Current assignee: NANJING AUDIT UNIVERSITY
Priority date: 2022-06-17
Filing date: 2022-06-17
Publication date: 2022-08-09

Abstract

The invention relates to the field of financial data storage, in particular to a financial data accurate tracing method based on a block chain MPT tree, which comprises the steps of firstly, carrying out different levels of granularity division on financial data based on a rough set theory; then combining the MPT tree with a multi-level multi-granularity structure model of the financial account data, carrying out hash operation on the financial data content on branch nodes and leaf nodes of the block chain MPT tree, and constructing the MPT tree of the financial account data so as to realize block chain MPT hierarchical structured storage of the financial data content; and finally, based on the block chain hash value and the timestamp, realizing accurate tracing of the financial data content. The distributed multi-centralization storage of the block chain has the advantage of tamper-resistant trusted storage; by adopting the method of fusing the structural Hash and the content Hash, the financial data can be accurately retrieved and traced based on the timestamp and the Hash value.

Description

Financial data accurate tracing method based on block chain MPT tree

Technical Field

The invention relates to the field of financial data storage, in particular to a financial data accurate tracing method based on a block chain MPT tree.

Background

At present, the existing financial data storage mainly adopts centralized storage, and the risk of potential tampering and untrusted storage exists; meanwhile, in the aspect of tracing, a log mode is mostly adopted for searching and tracing, and the accurate credible tracing of the content is lacked.

The block chain is a distributed storage architecture combining technologies such as point-to-point transmission, a consensus mechanism and a Hash algorithm, and has the characteristics of decentralization, anonymity, non-tamper property, traceability and the like. The method can realize credible storage and traceability of the financial data, but an accurate traceability technology of the hierarchical structured financial data is not available at present.

Disclosure of Invention

In order to solve the problems, the invention provides a financial data accurate tracing method based on a block chain MPT tree.

In order to achieve the purpose, the invention adopts the technical scheme that:

a financial data accurate tracing method based on a block chain MPT tree is characterized by comprising the following steps: the method comprises the following steps of performing multilevel and multi-granularity decomposition on bill record information based on a rough set theory of a set theory method, and realizing accurate multilevel and multi-granularity credible storage of the bill record information through an MPT (message passing test) tree of a block chain and a corresponding hash value technology, and specifically comprises the following steps:

s1 multilevel multi-granularity decomposition of financial account data based on rough set

S11, firstly, introducing a rough set theory into the construction of a grain structure model of financial bill data, defining a set of each piece of data in the balance sheet with a domain U, where an attribute set a is { a 1: primary subject name, a 2: name of secondary subject, a 3: third-level subject names }, f is a corresponding relation between each piece of data and attributes, namely, the subject names of the levels to which the data belong, and the value ranges of the attributes are respectively: va1 ═ asset, liability, owner's equity, Va2 ═ flowing asset, non-flowing asset, flowing liability, non-flowing liability, real income capital, capital equity, earnings, unallocated profit, Va3 ═ monetary capital, accounts receivable, inventory, accounts payable, accounts receivable … };

s12, defining three different equivalence classes according to the order of the granularity from small to large: 1) b ═ a1, a2, a3 in the unrecognizable relationship r (B), that is, data in the balance sheet are divided based on the names of the first, second and third-level subjects, and the corresponding data under each third-level subject is an equivalence class (particle); 2) b ═ a1, a2}, namely, data in the balance sheet are divided according to the names of the primary and secondary subjects, and the corresponding data under each secondary subject is an equivalence class (particle); 3) b ═ a1, namely dividing data in the balance sheet according to the names of primary subjects, wherein the corresponding data under each primary subject is an equivalent class (particle);

s13, dividing the data in the balance sheet into different types according to the granularity to form different grain layers, and combining the different grain layers to construct a multi-level and multi-granularity structural model of the financial account data;

s2, multi-level and multi-granularity precision trusted storage of financial account data based on block chain MPT tree and hash value

Combining the MPT tree with a multilevel multi-granularity structural model of the financial account data to construct the MPT tree of the financial account data, wherein the key of each node represents an enterprise, a year, a subject code and an amount code, and the values of the expansion nodes and the leaf nodes store corresponding amounts. The bottom layer of the tree is the hash value of the leaf node, hash values on the same branch are subjected to hash operation, and the like, and finally the root hash of the MPT tree is generated;

s3, based on the block chain hash value and the timestamp, accurate tracing of the financial data content is achieved;

s31 financial bill information tracing based on time stamp

Respectively storing the information types, the information contents and the MPT root hash in the original voucher, the bookkeeping voucher and the financial statement of the financial bill in a block chain, and adding a time stamp to record the recording time of the information in each block, wherein different blocks are connected according to the time sequence, and the time stamps are in one-to-one correspondence with the contents in the blocks;

then, the information category and the information content in the block are traced based on the timestamp, and the original voucher is positioned and traced through the timestamp, so that the original voucher can be arranged in time sequence, and the coming and going pulses of all economic services are proved;

s32 financial bill information tracing based on block hash

Adding the hash of the last block and the hash of the current block into each block of a financial bill information block chain, connecting the blocks based on the block hashes, and encoding a time stamp, an MPT root hash, a bill information category, bill information content and the like in the blocks through the block hashes to ensure the integrity and the non-tamper-ability of the content in the blocks;

the information type and information content in the block are then traced based on the block hash.

The invention has the following beneficial effects:

(1) compared with the traditional centralized storage, the distributed multi-centralized storage of the block chain has the advantage of tamper-resistant trusted storage;

(2) compared with the defect that the information content in the current block chain MPT tree is stored in leaf nodes more and cannot reveal the hierarchy and multi-granularity of the information content, the method for fusing the structural Hash and the content Hash can accurately retrieve and trace the financial data based on the timestamp and the Hash value.

Drawings

Other features, objects and advantages of the invention will become more apparent upon reading of the detailed description of non-limiting embodiments with reference to the following drawings:

FIG. 1 is a financial bill granule structure model.

Fig. 2 is a mobile bill data MPT tree data section.

Fig. 3 is the mobile bill data MPT tree content hash section.

Fig. 4 is a mobile bill data MPT tree structure hash section.

Fig. 5 is a block diagram of a timestamp based billing information block chain.

Fig. 6 is a block chain diagram of bill information based on block hash.

Fig. 7 is a financial billing information security traceability system.

FIG. 8 is a block chain platform architecture

FIG. 9 Mobile Bill data input interface

In the figure: (a) user a personal billing data; (b) user B package and flat fee data.

FIG. 10 a mobile billing data query interface;

in the figure: (a) a charging period of the user A; (b) the user A consumes in the current period; (c) and the user B calls the reminding service.

Detailed Description

The present invention will be described in detail with reference to specific examples. The following examples will assist those skilled in the art in further understanding the invention, but are not intended to limit the invention in any way. It should be noted that variations and modifications can be made by persons skilled in the art without departing from the spirit of the invention. All falling within the scope of the present invention.

The financial data accurate tracing method based on the block chain MPT tree in the embodiment of the invention is characterized in that a rough set theory based on a set theory method carries out multi-level and multi-granularity decomposition on bill record information, and the bill record information is accurately stored in a multi-level and multi-granularity credible mode through the block chain MPT tree and a corresponding hash value technology, and specifically comprises the following steps:

(1) multi-level multi-granularity decomposition of financial account data based on rough set

The particle is the most basic element in a granularity analysis model, the PAWLAK provides a particle representation method in a rough set model based on a set theory method, and an information system in a research problem is defined as a triple: s ═ U, a, f, where U is the domain of discourse, a is the set of attributes, and there is f: u × a → V, where V ═ Va, Va is the range of attribute a ∈ a. On the basis of the above-mentioned technical scheme,

existence of an unidentifiable relationship R (B):

U/R(B)＝{[x]b | x ∈ U } represents all equivalence classes derived from R on U, i.e., all grains on U.

The rough set theory is introduced into the construction of a grain structure model of financial bill data, a set of each piece of data in an asset liability table is defined by a domain U, and an attribute set A is { a 1: primary subject name, a 2: name of secondary subject, a 3: third-level subject names }, f is a corresponding relation between each piece of data and attributes, namely, the subject names of the levels to which the data belong, and the value ranges of the attributes are respectively: va1 ═ asset, liability, owner's equity, Va2 ═ flowing asset, non-flowing asset, flowing liability, non-flowing liability, real income capital, capital equity, earnings, unallocated profit, Va3 ═ monetary capital, accounts receivable, inventory, accounts payable ….

Three different equivalence classes are defined next in order of the particle size from small to large: 1) b ═ a1, a2, a3 in the unrecognizable relationship r (B), i.e. data in the balance sheet are divided based on the names of the first, second and third-level subjects, and the corresponding data under each third-level subject is an equivalence class (particle). In this case, the data has the smallest granularity, the highest refinement, but the lowest integration. Meanwhile, the more detailed account set differences among different enterprises are larger, so that the comparability of the data is the lowest; 2) b ═ a1, a2, i.e. data in the balance sheet are divided according to the names of the primary and secondary subjects, and the corresponding data under each secondary subject is an equivalence class (particle). The granularity of the data is moderate in the condition, and the refinement degree, the comprehensive degree and the comparability of the data are moderate; 3) b ═ a1, that is, data in the balance sheet is divided according to the names of the primary subjects, and the corresponding data under each primary subject is an equivalent class (particle). In this case, the granularity of the data is the largest, the refinement of the data is the lowest, but the comprehension and comparability are the highest.

After the data in the balance sheet are divided into different types according to the granularity, different grain layers are formed, and the different grain layers are combined to construct a multi-layer grain structure model as shown in fig. 1 (in order to save storage space, the names of the tables and the subjects are coded by using English letters). By constructing the multilayer grain structure model, the refinement degree of the data is improved, the comprehensive degree and comparability of the data are ensured, and the quality of information contained in the balance sheet is improved.

(2) Financial account data multilevel multi-granularity precision trusted storage based on block chain MPT tree and hash value

The hash algorithm is a function for mapping plaintext with any length into encrypted information with fixed length, two plaintext sections with different contents are difficult to find, and hash values of the two plaintext sections are consistent, so that the hash value has collision resistance, and the hash algorithm is mainly used for guaranteeing the safety of the information and improving the utilization rate of a data storage space. The MPT tree is a tree-shaped data structure established by a hash algorithm and is used for storing transaction information, states and corresponding state changes of accounts in a blockchain platform such as an Ethern. In the MPT tree, several hash values of the same branch are subjected to hash operation to generate a 'sub-hash', so that a smaller number of new first-level hashes are obtained, and the operations are analogized upwards in this way, and finally an inverted tree is formed. The specific structure of the MPT tree mainly comprises three nodes: leaf nodes (leaf nodes), extension nodes (extension nodes), and branch nodes (branch nodes), the key of each type of node representing the real path from the tree root to the node, the values of the extension and leaf nodes being the stored data, and the value of the branch node representing the node pointed to. In addition, a prefix (prefix) is contained in front of each of the extension node and the leaf node and is used for distinguishing the parity of the two nodes and the node key. The MPT tree carries out accurate and reliable storage on multi-level and multi-granularity bill record information with correlation in a layered Hash operation mode, finally generates a root Hash, and can safely store and verify data in the whole MPT tree through the root Hash.

And combining the MPT tree with the multi-layer structure of the mobile bill data to construct the MPT tree of the mobile bill data. Taking users a and B as examples, fig. 2 is an actual data part of the MPT tree of mobile bill data, a key of each node represents a user name, an information category and a project name, and a value of a leaf node stores corresponding information. Fig. 3 shows the content hash part of the MPT tree for mobile bill data, where each hash value corresponds to an actual piece of data stored in the MPT tree. Fig. 4 is a structural hash part of the MPT tree for mobile bill data, where the bottom layer of the tree is the hash value of a leaf node, hash values on the same branch are subjected to hash operation, and so on, and finally the root hash of the MPT tree is generated. Fig. 2, 3 and 4 collectively store 8 billing data of users a and B, as shown in table 1.

And combining the MPT tree with a multi-level and multi-granularity structural model of the bill record information to construct the MPT tree of the financial bill data, wherein the MPT tree is used for storing the bill report data of the enterprise. Taking enterprise a as an example, fig. 5 is an actual data part of the MPT tree of the enterprise a bill statement data, a key of each node represents an enterprise, a year, a subject code, and an amount code, and values of the extended node and the leaf node store corresponding amounts. Fig. 6 is a structural hash part of the MPT tree of the enterprise a bill report data, the bottom layer of the tree is the hash value of the leaf node, hash values on the same branch are subjected to hash operation, and so on, and finally the root hash of the MPT tree is generated. Fig. 5 and 6 collectively store 8 pieces of bill report data of the enterprise a, and the corresponding enterprise name, report year, report name, subject name, and amount are shown in table 1.

TABLE 1 Mobile Bill data for Users A and B

(3) Financial billing information tracing based on timestamps

Each block of the block chain comprises a time stamp indicating the writing time of the block data, and the time stamp system can be used as the existence certification of the block data by virtue of the legal authority and authorization status of the time stamp system, so as to achieve the purposes of non-repudiation and non-repudiation. Therefore, the blocks in the block chain are arranged in sequence according to the time sequence, so that things happening at a certain time in history can be traced and verified from the block chain. The information category, the information content and the MPT root hash in the original voucher, the bookkeeping voucher and the financial statement of the financial bill are respectively stored in a block chain, and the entry time of the information is stamped in each block, as shown in fig. 5. Different blocks are connected according to the time sequence, and the time stamps are in one-to-one correspondence with the contents in the blocks.

Then, the information category and the information content in the block are traced based on the timestamp, and the original voucher is positioned and traced through the timestamp, so that the original voucher can be arranged according to a time sequence, and the coming and going of all economic services are proved; the accounting vouchers are positioned and traced through the timestamps, which is equivalent to automatically generating numbers and compiling dates in the accounting vouchers, so that the ordering of the accounting vouchers is facilitated; the data in the bill report is positioned and traced through the timestamp, so that the time dimension can be added for each piece of data, and the source of each piece of data is determined.

(4) Financial billing information tracing based on block hashing

The block hash is a digital signature of a block generated by a hash algorithm, and each block of the block chain includes the hash of the last block and the hash of the current block. If the content of the previous block or the current block is changed, the hash of the current block is changed, and the block originally following the current block cannot be linked to the current block, which means that the block chain is broken, so that the block information cannot be artificially tampered and forged. Here, the hash of the last block and the hash of the current block are added to each block of the fiscal bill information block chain, and the blocks are connected based on the block hashes, as shown in fig. 6. And encoding the time stamp, the MPT root hash, the bill information category, the bill information content and the like in the block through the block hash, so that the integrity and the non-tamper property of the content in the block are ensured.

Then, the information category and the information content in the block are traced based on the block hash, and the bill information in the original voucher is traced through the block hash, so that each bill service can be known exactly, all information in the voucher is verified, and the information is used as the basis for auditing the bookkeeping voucher; tracing the bill information in the bookkeeping voucher, verifying whether the content of the bill information is consistent with the original voucher, whether the listed accounting subjects, lending directions and amounts are correct, whether the information such as dates and summaries is completely filled, and confirming related responsible persons such as an accounting supervisor and bookkeeping personnel; the method has the advantages that the billing information in the financial statement is traced, the tracing and the query of the single data corresponding to the specific subject are facilitated, and the method is more accurate and efficient compared with the tracing and the query of the whole statement file.

A timestamp-based tracing manner and a block hash-based tracing manner are combined to construct a bill information security tracing system, as shown in fig. 7. And the enterprise financial department writes the financial information such as the original voucher, the bookkeeping voucher, the financial statement and the like and uploads the financial information to the block chain network, the relevant departments in the enterprise, the government auditing organ and the social auditing mechanism send a request to the block chain tracing system by using the stored timestamp and the current block hash as the tracing code, and the system background is automatically positioned to the corresponding block to obtain the bill information category and the bill information content. Financial information is provided for relevant auditing agencies in this way, and if any account problem occurs, the problem source can be found rapidly and pertinently accountability is performed.

Examples

The block chain platform of enterprise financial information management is set up in this embodiment simulation, realizes accurate location and the storage of financial statement data to and the credible of relevant financial information is traceed back, then simulates and snatchs relevant data, verifies the model through the experiment.

(1) Block chain platform construction

A Go language is used for simulating and building a block chain platform in a Linux system, the main architecture is as shown in FIG. 8, and a data layer mainly stores related information through a block and an MPT tree; in the network layer, a distributed network consisting Of 6 nodes is included, any nodes are connected with each other and can carry out information transmission, and all the nodes achieve the information consistency through a POW (proof Of office) consensus mechanism; the application layer comprises accurate storage and query of financial statement data and tracing of financial information.

The main functions implemented by the platform include the following 4 parts:

1) and (3) generation of a block: the financial staff selects the entered financial information category (original voucher/bookkeeping voucher/financial statement) and then fills in the corresponding financial information content. And when information is recorded once, the system background automatically operates a POW consensus mechanism and generates a new block, the new block is connected with the existing block chain to form a new block chain, the block chain information of all the nodes is updated, and finally the new block chain is stored in the BoltDB database after being serialized.

2) Updating the MPT tree: and according to the financial statement information input by financial staff, taking the enterprise name, the statement year and the subject code as keys, and inserting the data which is subjected to RLP coding as value into the existing MPT tree. And the new branch and the original MPT tree are subjected to Hash calculation to generate new root Hash, so that the MPT tree is updated. In addition, the MPT trees are stored in a levelDB database through serialization processing.

3) Query of data in MPT tree: the method comprises the steps of firstly inputting an MPT root stored in a block, reading an existing MPT tree in a levelDB database, then inputting an enterprise name, a report year and a subject code, and automatically inquiring corresponding financial report data in the MPT tree by taking the input information as a key by a system background.

4) Tracing the block information: firstly, reading an existing block chain in a BoltDB database, then inputting a tracing code of a block to be traced, namely a timestamp or block hash, automatically positioning a system background to the corresponding block, and returning the corresponding financial information type and the financial information.

(2) Example verification

The above 8 pieces of mobile bill data are actually stored in the MPT tree by Go language, and the information input interface is as shown in fig. 9.

The data entered in the MPT tree is queried, and the query interface is shown in fig. 10.

The foregoing description of specific embodiments of the present invention has been presented. It is to be understood that the present invention is not limited to the specific embodiments described above, and that various changes and modifications may be made by one skilled in the art within the scope of the appended claims without departing from the spirit of the invention.

Claims

1. A financial data accurate tracing method based on a block chain MPT tree is characterized by comprising the following steps: the method is characterized in that multilevel and multi-granularity decomposition is carried out on bill record information based on a rough set theory of a set theory method, and accurate multilevel and multi-granularity credible storage of the bill record information is realized through an MPT tree of a block chain and a corresponding hash value technology.

2. The method for accurately tracing financial data based on the block chain MPT tree as claimed in claim 1, wherein: the method comprises the following steps:

Combining the MPT tree with a multi-level multi-granularity structural model of the financial account data to construct the MPT tree of the financial account data, wherein the key of each node represents an enterprise, a year, a subject code and a money amount code, the values of the expansion nodes and the leaf nodes store corresponding money amounts, the bottom layer of the tree is the hash value of the leaf node, hash values on the same branch are subjected to hash operation, and the like, and finally the root hash of the MPT tree is generated;

s31 financial bill information tracing based on time stamp

s32 financial bill information tracing based on block hash

Adding the hash of the last block and the hash of the current block into each block of the financial bill information block chain, connecting the blocks based on the block hashes, and coding a time stamp, an MPT root hash, a bill information category and bill information content in the block through the block hashes to ensure the integrity and the non-tamper property of the content in the block;