An intelligent auditing system based on blockchain technology
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- HAO JING COLLEGE OF SHAANXI UNIV OF SCI & TECH
- Filing Date
- 2026-02-13
- Publication Date
- 2026-06-05
Smart Images
Figure CN122155638A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the fields of audit informatization and blockchain technology, and in particular to an intelligent audit system based on blockchain technology. Background Technology
[0002] As enterprises become increasingly digitalized, audit targets are gradually shifting from traditional paper-based vouchers to electronic, systematic, and multi-source heterogeneous data formats. Existing audit systems typically rely on data provided by internal enterprise information systems or third-party platforms for audit analysis. However, in practice, problems such as difficulty in quantifying the authenticity of data sources, the ease with which audit data can be tampered with after the fact, the difficulty in retrospectively verifying audit conclusions due to changes in audit rules over time, and unclear attribution of audit responsibility are still prevalent.
[0003] On the one hand, traditional auditing systems often simply categorize data as "credible or unreliable," lacking the technical means to quantitatively assess the reliability of data sources, making it difficult to provide refined basis for automated audit analysis. On the other hand, existing systems mostly record manual judgments and reviews generated during the audit process in log form, lacking anti-tampering and accountability mechanisms, making it difficult to meet the traceability requirements in regulatory and judicial scenarios. Furthermore, auditing standards and rules have dynamic evolutionary characteristics, and existing auditing systems generally lack rule version management and historical replay capabilities, making it difficult to interpret or verify audit conclusions afterward.
[0004] Therefore, there is an urgent need for a new technical solution that can ensure the authenticity and integrity of audit data, while also enabling versioned management of audit rules, traceable and solidified audit responsibilities, and verifiable audit results, so as to improve the intelligence and credibility of the audit system. Summary of the Invention
[0005] This invention addresses the technical problems existing in the background art by proposing an intelligent auditing system based on blockchain technology.
[0006] To solve the technical problem, the technical solution of the present invention is as follows:
[0007] A smart auditing system based on blockchain technology, the system comprising: The trusted data collection and scoring module is used to connect to data sources and collect data to be audited; calculate the hash value of the data to be audited, and generate the corresponding trusted assessment information based on multiple preset trusted assessment dimensions; and send the trusted data packet including the hash value, timestamp and trusted assessment information of the data to be audited to the blockchain evidence storage and responsibility chain module. The blockchain evidence storage and responsibility chain module is equipped with smart contracts to receive and store the trusted data packets, forming a tamper-proof audit data chain. When performing key audit operations, it records the information of the first responsible party initiating the audit operation and the information of the second responsible party confirming the audit operation, forming a responsibility chain record associated with the audit data. The versioned intelligent audit engine module stores versioned audit rules with version identifiers that are effective within a specified time interval. This module is configured to: obtain the trusted data packet from the blockchain evidence storage and responsibility chain module, call the effective versioned audit rules corresponding to the current audit time point, and automatically analyze the data to be audited in conjunction with the credibility assessment information to generate versioned audit clues. The human-machine collaborative auditing module is used to receive the versioned audit clues, provide auditors with an interactive interface to execute in-depth auditing procedures based on the versioned audit clues, and submit the manual audit evidence and manual audit judgments generated by the auditors in the interactive interface to the blockchain evidence storage and responsibility chain module for evidence storage, so as to update the audit data chain and the responsibility chain records. The audit report and verification service module is used to generate a final audit report based on the complete audit trajectory stored in the blockchain evidence storage and responsibility chain module, which includes the trusted data packet, the versioned audit clues, the manual audit evidence and the manual audit judgment, and the responsibility chain record, and to calculate and publish the summary hash value of the final audit report.
[0008] Furthermore, the preset multiple credibility assessment dimensions include: the authority level of the data source, the type of data collection interface, the degree of redundancy of the data acquisition method across multiple nodes, and the historical accuracy of the data provided by the data source.
[0009] Furthermore, the smart contracts deployed in the blockchain evidence storage and responsibility chain module specifically include an evidence storage contract and a responsibility chain contract; the evidence storage contract is used to verify and store the trusted data packet; the responsibility chain contract is used to generate a responsibility chain record when the human-machine collaborative auditing module submits the manual audit evidence and judgment, and the responsibility chain record includes: an operation type identifier, a digital identity identifier of the responsible entity of the corresponding auditor, a hash value or clue identifier of the original record, and multi-party digital signatures.
[0010] Furthermore, the versioned audit trail includes: the target rule version identifier that triggers the generation of the versioned audit trail, the hash value set of the audit data involved, the data hash value set, and the credibility assessment information summary.
[0011] Furthermore, the versioned intelligent audit engine module is configured as follows: Based on the timestamp in the trusted data packet and the effective time interval of the versioned audit rule, determine the target rule version identifier that should be invoked at the current audit time point; During the automated analysis process, a differentiated verification strategy for the data to be audited is dynamically generated and executed based on the credibility assessment information. The target rule version identifier, the automated audit findings generated after executing the differentiated verification strategy, and the set of hash values of the audit data involved are used together as the versioned audit clues.
[0012] Furthermore, the human-machine collaborative auditing module is further configured as follows: Upon receiving the versioned audit trail, based on the credibility assessment information summary contained in the versioned audit trail, a list of associated differentiated audit tasks is dynamically organized and presented to the auditors. The list of differentiated audit tasks includes the data range to be reviewed, the suggested audit procedure types, and priorities. The auditors perform the differentiated audit tasks through the interactive interface and submit corresponding manual audit evidence and manual audit judgments.
[0013] Furthermore, the audit report and verification service module is further configured as follows: When generating the final audit report, the versioned audit clues, manual audit judgments and corresponding responsibility chain records associated with the complete audit trajectory are structurally linked to generate an embedded evidence index with verifiable pointing relationships.
[0014] Furthermore, the audit report and verification service module is also equipped with a verification interface; The verification interface is configured to: receive a report to be verified submitted by a user, and by parsing the embedded evidence index, retrieve and compare the hash value of the corresponding original record stored in the blockchain evidence storage and responsibility chain module to verify the integrity and tamper-proof nature of the report to be verified.
[0015] Furthermore, the versioned intelligent audit engine module is also configured to: respond to a historical replay instruction, based on a specified historical time point, invoke the versioned audit rules that are in effect at the historical time point and the trusted data snapshot based on the historical time point, recalculate the historical audit conclusions, and generate historical replay audit results for comparison and analysis. The historical replay audit results are stored through the blockchain evidence storage and responsibility chain module and used for comparison and analysis.
[0016] Furthermore, the audit report and verification service module is further configured as follows: Based on the credibility assessment information, the certainty level of the versioned audit rules, and the review level of the manual audit judgment, the evidence strength level corresponding to the audit conclusion is calculated, and when generating the final audit report, the evidence strength level is marked and embedded into the evidence index of the report.
[0017] Furthermore, the calculation of the strength level of evidence includes: The data credibility score is determined based on the credibility assessment information; the rule reliability score is determined based on the rule certainty level of the versioned audit rules; the manual review score is determined based on the number of review levels experienced by the manual audit judgment; the above scores are combined according to the preset combination rules to generate the corresponding evidence strength level, and marked in the final audit report.
[0018] This application has the following advantages: First, by setting up a trusted data collection and scoring module, this invention introduces a multi-dimensional credibility assessment mechanism to quantitatively evaluate the authority of the data source, collection method, and historical accuracy of the data to be audited, avoiding the simplistic assumption that all data is equally credible, thus improving the accuracy and reliability of automated audit analysis from a technical perspective.
[0019] Secondly, this invention utilizes blockchain and smart contract technology to immutably preserve audit data, audit clues, and manual auditing actions. It also uses a responsibility chain contract to record the primary and secondary responsible parties for key audit operations in a chain, thereby achieving automatic solidification and traceability of audit responsibility and effectively preventing disputes over responsibility and denial of audit results.
[0020] Furthermore, by introducing versioned audit rules and a matching mechanism for rule effective time intervals, this invention ensures that each audit conclusion can be clearly traced back to its corresponding data version and rule version, and supports the replay and comparative analysis of historical audit conclusions, thereby significantly improving the interpretability and time consistency of audit conclusions.
[0021] Furthermore, this invention combines automated auditing with professional human judgment through a human-machine collaborative auditing module, and embeds verifiable evidence indexes and evidence strength level markings in the audit report, thereby further enhancing the transparency, credibility, and regulatory applicability of the audit report. Attached Figure Description
[0022] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0023] Figure 1 This application provides a technical roadmap for an intelligent auditing system based on blockchain technology. Detailed Implementation
[0024] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.
[0025] Example 1: A smart auditing system based on blockchain technology, such as Figure 1 As shown, the system includes: The trusted data collection and scoring module is used to connect to data sources and collect data to be audited; calculate the hash value of the data to be audited, and generate the corresponding trusted assessment information based on multiple preset trusted assessment dimensions; and send the trusted data packet including the hash value, timestamp and trusted assessment information of the data to be audited to the blockchain evidence storage and responsibility chain module. The Trusted Data Acquisition and Scoring module serves as the starting point for the entire audit process, undertaking the crucial task of extracting and initially evaluating data from diverse real-world data. Its work is not simply data transfer, but a sophisticated processing procedure comprising three core stages, designed to deliver semantically consistent, clearly identified, and inherently trust-assessed standardized data packages to all subsequent stages.
[0026] First, the module connects to various heterogeneous data sources, such as business databases, API interfaces, or file systems, through configurable adapters. After collecting raw data, its core task is to perform standardized collection and encapsulation, that is, to perform unified format cleaning, transformation, and semantic alignment, converting the raw data into a structured set of auditable data records. This step ensures that data from different sources has consistent meaning and format within the system, laying a reliable foundation for subsequent automated analysis.
[0027] Subsequently, the module generates an immutable digital identity for each piece of standardized data. It uses a cryptographic hash function to calculate a unique hash value for the data, serving as its digital fingerprint. Simultaneously, the module obtains a precise timestamp from a trusted time source and binds it to the hash value, together accurately recording the moment the data was captured by the system. This is equivalent to issuing a birth certificate with a unique number and precise time for each piece of data.
[0028] After identity verification, the module further quantifies and evaluates the inherent credibility of the data. Based on a pre-defined multi-dimensional model, it automatically analyzes the data's source and quality and generates a score. Evaluation dimensions typically cover multiple aspects, including the data source's authority, the security of the data collection interface, redundancy and consistency with data from other sources, and the historical accuracy of the data provided by the data source. Finally, the module packages the data's hash fingerprint, timestamp, and this structured credibility assessment information into a complete trusted data packet. This data packet is then sent to the blockchain for permanent storage, thus laying the first solid and verifiable foundation of trust for the entire audit process.
[0029] Inputs: data source identifier sid, interface type api_type, number of collection nodes k, multi-node consistency, historical accuracy acc, whether TLS / signature verification sig_ok, etc. Output: Dimensional scores S_auth, S_api, S_redun, S_hist; Total score S_total; Level L∈{high / medium / low}; Automated analysis process: Parse the data packet metadata: source ID, collection channel, collection node, signature verification result, timestamp, etc.; Get S_auth by calling the authority mapping table; S_api is mapped according to the interface type (e.g., Bank Direct Connection API > ERP Read-Only Interface > File Import). Redundancy consistency score: S_redun = floor(consistency) 100), consistency = number of consistent nodes / k; Historical accuracy score: S_hist = floor(acc) 100) (acc is calculated by inversely estimating the historical reconciliation / sampling error rate). In summary: S_total = Σ(w_i S_i); Grading: S_total≥80 High; 50-79 Medium; <50 Low.
[0030] Furthermore, the trusted data acquisition and scoring module is further configured as follows: Based on a preset credibility assessment model, the data to be audited is quantitatively scored across multiple credibility assessment dimensions, and structured credibility assessment information is generated. The credibility assessment model includes: configuring corresponding scoring rules and weight parameters for each credibility assessment dimension; mapping the authority level of the data source, the type of data collection interface, the multi-node redundancy of the data acquisition method, and the historical accuracy of the data source to standardized dimension score values; generating a comprehensive credibility score for the audited data using a weighted calculation method based on the score values of each dimension and the corresponding weight parameters; the credibility assessment information includes: the score values of each credibility assessment dimension, the comprehensive credibility score, and the corresponding credibility level identifier; the credibility level identifier is used to trigger differentiated analysis strategies and audit task arrangement in subsequent audit processes.
[0031] The scoring rules and weighting parameters include: Authority level: A / B / C / D → 100 / 80 / 60 / 40; Interface type: Dedicated line API / HTTPS + signature / Internal read-only interface / File import → 100 / 85 / 70 / 50; Redundancy level: k≥3 and consistency rate≥0.95 → 100; consistency rate 0.8-0.95 → 80; <0.8 → 50; Historical accuracy: acc ≥ 0.99 → 100; 0.95 - 0.99 → 80; < 0.95 → 60; Weights: w_auth=0.35, w_api=0.25, w_redun=0.20, w_hist=0.20.
[0032] Furthermore, the trusted data packet only contains the hash value of the data to be audited, while the original content of the data to be audited is stored in an off-chain data storage system. The off-chain data storage system is configured to: index and store the original data according to the hash value; when a data retrieval or verification request is received, the hash value of the corresponding original data is recalculated based on the hash value and compared with the hash value stored in the blockchain; when the comparison result is consistent, it is confirmed that the original data has not been tampered with; when the comparison result is inconsistent, it is determined that the data integrity verification has failed; through the above dual verification mechanism of on-chain hash value and off-chain original data, the technical guarantee of the integrity and consistency of the audited data is achieved.
[0033] The blockchain evidence storage and responsibility chain module is equipped with smart contracts to receive and store the trusted data packets, forming a tamper-proof audit data chain. When performing key audit operations, it records the information of the first responsible party initiating the audit operation and the information of the second responsible party confirming the audit operation, forming a responsibility chain record associated with the audit data. EvidenceStore, a contract for storing evidence. storeEvidence(hash_data, ts, score_digest, sender_did): Verify the sender's DID certificate → write to the on-chain mapping hash_data ->EvidenceMeta{ts, score_digest, sender_did}; ResponsibilityChain; Structure: Record{opType, targetHash, firstDid, secondDid, sig1, sig2, status, deadline}; initRecord(opType, targetHash, firstDid, sig1): Verify signature sig1 → Write status=PENDING_CONFIRM; confirmRecord(recordId, secondDid, sig2): Verify signature sig2→status=CONFIRMED; Timeout: if now > deadline && sig2 missing => status = UNCONFIRMED; Signature verification method: Based on ECDSA, the signature is verified using the message digest of (opType|targetHash|ts).
[0034] Specifically, the blockchain-based evidence storage and accountability chain module serves as the system's core trust anchor and process traceability hub. Through smart contracts deployed on the blockchain, it achieves two core functions: first, it solidifies front-end data packets into an immutable audit trail, i.e., the audit data chain; second, it binds clear operational responsibilities to key human decision-making stages, i.e., the accountability chain record, thereby achieving end-to-end data trustworthiness and operational traceability. Internally, it consists of two core smart contracts and an on-chain data structure working collaboratively, with the specific logic as follows: The blockchain-based evidence storage and accountability module serves as the core trust anchor of the system. Through smart contracts deployed on the chain, it transforms key data and human operations in the audit workflow into immutable and clearly defined trustworthy records. Its internal operating logic and the final on-chain data structure together form the cornerstone of the system's traceability and verifiability.
[0035] The core of this module is implemented collaboratively by two clearly defined smart contracts. The evidence storage contract is primarily responsible for solidifying data. Its function is to receive and permanently store trusted data packets from the front end, establishing an initial anchor for audit data. The specific workflow is as follows: First, the identity of the data packet sender is verified, and the data packet is received after confirmation. Subsequently, the contract calculates the hash value of the data packet itself and writes it, along with the timestamp, sender identity, and other information, into the blockchain, generating a tamper-proof data anchor. The core value of this anchor lies in the fact that it points to the original data fingerprint calculated by the front end through the hash value, thereby forming an audit data chain on the chain that can be verified step by step.
[0036] The responsibility chain contract focuses on binding responsibility. Its function is to create an undeniable record of responsibility at key operational nodes requiring human intervention. When auditors perform such operations, the contract is triggered, recording the operation type, operation objective, and, most importantly, the information of the two responsible parties—the digital identities of the auditor initiating the operation and the person verifying it. The contract requires and verifies the digital signatures of both parties on the operation record to ensure their awareness and approval, and finally uploads the complete signature record to the blockchain, forming an operational anchor.
[0037] Furthermore, the responsibility chain contract is further configured to: trigger the responsibility chain record generation process upon detecting a preset key audit operation type, wherein the key audit operation type includes at least: clue confirmation, audit judgment submission, and audit conclusion review; before generating the responsibility chain record, the responsibility chain contract verifies the digital signature submitted by the operation initiator to verify the legitimacy of the first responsible party's identity; in scenarios requiring confirmation from a second responsible party, the responsibility chain contract is configured to: enter a pending confirmation state after the first responsible party completes the signature; after receiving and verifying the digital signature of the second responsible party, generate a complete responsibility chain record and write it to the blockchain; if confirmation from the second responsible party is not obtained within a preset time, the responsibility chain record remains in an incomplete state.
[0038] The continuous operation of these two contracts ultimately weaves together a verifiable data structure on the blockchain. The audit data chain consists of a series of data anchors connected in chronological order, much like a birth certificate and fingerprint archive for data; the accountability chain record consists of a series of operational anchors, like precise thumbtacks pinned to specific data chain or clue nodes, clearly recording who made what judgment on which data at what time, and who confirmed it. In short, this module uses cryptographic technology to transform data states and human operations into a series of time-ordered, interconnected, and multi-signature trusted records, laying the foundation for a clear, traceable, and verifiable trust basis for the entire system.
[0039] The versioned intelligent audit engine module stores versioned audit rules with version identifiers that are effective within a specified time interval. This module is configured to: obtain trusted data packets from the blockchain evidence storage and responsibility chain module, call the effective versioned audit rules corresponding to the current audit time point, and automatically analyze the data to be audited in conjunction with the credibility assessment information to generate versioned audit clues. Specifically, the versioned intelligent audit engine module is the core of the system's automated analysis, and its core innovation lies in the introduction of time-dimensional rule management. By executing audit rule versions that are strictly matched to specific points in time, it ensures that all automated analysis conclusions have clear, traceable, and non-repudiable logical basis.
[0040] The core innovation of the versioned intelligent audit engine module lies in its deep integration of audit rules with the time dimension and its dynamic strategy adjustment through credibility assessment, thereby driving automated analysis. This module does not simply execute a set of static rules, but is an intelligent system capable of understanding the spatiotemporal context and data credibility. Its operation relies on the close collaboration of two core components. First, the versioned rule base acts as a legal code. All audit rules in the base, like legal provisions, have unique version identifiers and clearly defined effective time ranges. Any update or revision of a rule must be published as a new version, while the old version is archived, thus forming a complete and tamper-proof history of rule evolution. This ensures that any historical analytical conclusion can be traced back to the precise rule text upon which it was based.
[0041] Furthermore, the dynamic differentiated verification strategy generation includes: dividing the data to be audited into at least high, medium, and low credibility data levels based on the comprehensive credibility score in the credibility assessment information; configuring a corresponding set of verification strategy parameters for different credibility data levels, wherein the verification strategy parameters include at least: the number of verification rules, the strictness of the verification threshold, and the range of data participating in the verification; and when performing automated analysis, the versioned intelligent audit engine module selects the corresponding set of verification strategy parameters based on the data level and executes the corresponding automated audit analysis process.
[0042] Secondly, there's the rule enforcement engine, a combination of judge and detective. When the engine works, it first retrieves trusted data packets containing timestamps and credibility assessment information from the blockchain. The processing flow embodies a refined intelligence: the first step is spatiotemporal alignment, precisely matching the then-current rule version in the rule base based on the data's creation time, ensuring historical consistency of the analysis logic. Building on this, the engine enters the dynamic policy generation phase, parsing the credibility score in the data packet and automatically adjusting the rigor of the analysis. For example, for data with low credibility, it uses a more complex and conservative set of verification rules; while for high-credibility data, it may employ a more efficient standard process. Finally, the engine performs analysis, conducting a deep scan of the corresponding off-chain raw data based on the matched rule version and the generated dynamic policy, completing tasks such as anomaly detection and correlation analysis.
[0043] Based on the above mechanism, this module does not produce raw alerts, but rather highly structured, versioned audit leads with built-in audit trails. Each lead is equivalent to a miniature analysis report, which clearly records: the version identifier of the target rule that triggered the lead, enabling logical tracing; the hash value set of all original data associated with the lead, ensuring the inseparability of conclusions and evidence; a detailed description of the automated audit findings; and a key credibility context summary, providing important risk background for subsequent manual judgment. This design makes the process and results of automated analysis themselves auditable and verifiable.
[0044] Furthermore, the versioned intelligent audit engine module is further configured to: index and store each versioned audit rule according to its effective time interval; upon receiving a trusted data packet, search for the target rule version in the versioned audit rules that is in the same effective time interval as the timestamp in the trusted data packet; when multiple rule versions meet the time matching condition, determine a unique target rule version identifier according to the preset rule priority or the most recently effective principle; and use the target rule version identifier as the rule basis for this automated analysis.
[0045] Furthermore, after the historical replay instruction is triggered, the versioned intelligent audit engine module is configured to: retrieve the off-chain data snapshot at the corresponding time point based on the hash value of the historical trusted data packet recorded in the blockchain evidence; simultaneously invoke the versioned audit rules that took effect at the historical time point to perform automated audit analysis on the off-chain data snapshot; compare and analyze the historical replay audit results with the original audit conclusions, and store the historical replay audit results as a new audit record.
[0046] The human-machine collaborative auditing module is used to receive the versioned audit clues, provide auditors with an interactive interface to execute in-depth auditing procedures based on the versioned audit clues, and submit the manual audit evidence and manual audit judgments generated by the auditors in the interactive interface to the blockchain evidence storage and responsibility chain module for evidence storage, so as to update the audit data chain and the responsibility chain records. Specifically, the human-machine collaborative audit module is the central hub for integrating intelligent system analysis with professional human judgment. It transforms standardized clues generated by the automated engine into guiding audit tasks with clearly defined responsibilities, and anchors the key operational results of auditors to the blockchain in a structured and accountable manner, completing a trustworthy closed loop from machine discovery to human verification.
[0047] The human-machine collaborative audit module serves as a hub connecting machine intelligence and human professional judgment. Its workflow revolves around the in-depth processing of a versioned audit lead, achieving a closed loop from automated alerts to authoritative audit conclusions. The entire process begins with a lead and ends with the on-chain recording of responsibility, demonstrating a high degree of integration between intelligent guidance and clear accountability.
[0048] When the module receives a versioned audit lead from the intelligent engine, it first initiates intelligent task orchestration internally. The system automatically parses the key information embedded in the lead, especially the data credibility assessment summary and the triggered rule version identifier. Based on these parsing results, the module does not issue uniform instructions, but can dynamically generate a differentiated audit task list. For example, for a lead marked as high-risk or originating from low-credibility data, the list will suggest expanding the scope of data review and explicitly recommend performing more in-depth substantive detail testing procedures, while marking it as a high-priority task. Conversely, for low-risk leads, it may only orchestrate standard-priority analytical review procedures. This orchestration process essentially translates the credibility of the data itself into specific, actionable audit action guidelines.
[0049] Intelligent task orchestration specifically includes: Input: risk_level, S_total, rule determinism level R_cert, transaction amount, anomaly type; Output: Task List = [{task Type, scope, priority, suggested Procedure}]; Example of arrangement rules: If S_total < 50 or risk_level = high: priority = high; Scope = Expanded to the same customer / same period / same business line; procedure = confirmation / walkthrough test / deadline test / third-party reconciliation; If S_total≥80 and the exception type is slight threshold deviation: priority = medium / low; procedure = analytical review + sampling inspection.
[0050] Subsequently, auditors enter the guided interactive work phase. The system presents auditors with a clearly defined interface, including an intelligently curated task list, complete clue context, and a secure window that allows direct access to raw data via hash values. This design ensures that auditors simultaneously understand what needs to be done, why it needs to be done, and what original evidence is being used, greatly improving the relevance and efficiency of the review. Auditors can perform checks within this interface and upload supplementary evidence, record findings, and form preliminary audit judgments through an integrated panel; the complete clue context includes: rule version, data hash, and credibility details.
[0051] After auditors complete their verification and submit their judgment, the system immediately initiates a structured encapsulation and accountability process. All manual output, including the judgment conclusion, the hash value set of supporting evidence, and the logical explanation, is automatically encapsulated into a standard, digitally signed manual audit judgment object. This object is not only structured in content, but more importantly, it is signed with the auditor's private key, ensuring the non-repudiation of the operation.
[0052] Ultimately, this crucial manual operation will be synchronized to the blockchain, completing the final anchoring of trust. The module will automatically invoke the responsibility chain contract on the blockchain to record this manual judgment submission as a key event. The generated responsibility chain record will clearly record the operation type, the digital identity of the submitter (the primary responsible party), and the clues and object hashes associated with the judgment. In rigorous scenarios requiring review, the system will also guide the judgment to the secondary responsible party for confirmation and signature, thereby achieving dual signatures for key audit conclusions. This permanently locks the clear personnel responsibility onto the tamper-proof blockchain, completing a fully trusted closed loop from machine discovery to manual confirmation and then to responsibility solidification.
[0053] The audit report and verification service module is used to generate a final audit report based on the complete audit trajectory stored in the blockchain evidence storage and responsibility chain module, which includes the trusted data packet, the versioned audit clues, the manual audit evidence and the manual audit judgment, and the responsibility chain record, and to calculate and publish the summary hash value of the final audit report.
[0054] Specifically, the audit report and verification service module serves as the system's trust crystallization and public verification portal. As the official reader and verifier of the audit trail blockchain, it aggregates, interprets, and generates a dynamic digital report with complete verifiability from the scattered, original, but tamper-proof records on the chain.
[0055] As the final output port of the system's trust value, the audit report and verification service module's core mission is to transform the original, decentralized records on the blockchain into an authoritative, readable, and self-verifying dynamic digital report. This module completes the final delivery from on-chain data to trusted conclusions through three sequentially connected processing stages.
[0056] First, the module initiates full-trajectory aggregation and correlation reconstruction. Upon inputting a unique identifier for an audit project, the module acts as a smart query client on the blockchain, automatically retrieving and aggregating all record fragments related to that project from the chain. This includes all initial trusted data packets, derived versioned audit leads, human judgments and evidence hashes generated for each lead, and the chain of responsibility records of the operators involved. More importantly, the system can intelligently identify the hash pointers embedded between these records, automatically reconstructing a clear and directed evidence graph and responsibility flow diagram—from raw data to automated leads, and then to human judgments and responsibility attribution—like piecing together a complete jigsaw puzzle, providing a structured panoramic view for report generation.
[0057] Building upon this foundation, the module enters the structured report generation and trust enhancement phase. It doesn't simply compile on-chain logs into a document; instead, it generates a new type of report with an embedded verifiable digital kernel. Each key conclusion in the report is accompanied by an evidence index embedded in a non-intrusive format, allowing readers to easily access the original on-chain evidence chain, view the rule version that triggered the conclusion, the data fingerprint on which it was based, and the identities of the responsible parties. Simultaneously, the system comprehensively analyzes the credibility of the data upon which the conclusions rely, the certainty of the rules, and the level of human review, automatically calculating and labeling the strength of evidence, providing a clear indication of the robustness of the conclusions. The final report is presented in a readable document format conforming to industry standards, but it internally encapsulates a complete digital trust kernel.
[0058] Ultimately, the module completes the public delivery of trust through a report fingerprint publishing and verification service. It calculates the hash value of the entire final report, publishing it as the report's unique digital fingerprint on public channels such as the blockchain. Simultaneously, the module provides a public verification interface. Any report recipient can use this interface to verify whether the file hash of their report matches the officially published fingerprint, thus ensuring the integrity and tamper-proof nature of the report content. Furthermore, recipients can utilize the evidence index embedded in the report to initiate a thorough query through this interface, independently verifying the complete chain of evidence and accountability behind any conclusion. This mechanism transforms each audit report from a static document into a publicly verifiable node of trust.
[0059] Furthermore, the recipient can utilize the embedded evidence index in the report to retrieve the original on-chain evidence records item by item through this interface for verification, achieving a penetrating verification from conclusion to source.
[0060] Furthermore, the preset multiple credibility assessment dimensions include: the authority level of the data source, the type of data collection interface, the degree of redundancy of the data acquisition method across multiple nodes, and the historical accuracy of the data provided by the data source.
[0061] Furthermore, the smart contracts deployed in the blockchain evidence storage and responsibility chain module specifically include an evidence storage contract and a responsibility chain contract; the evidence storage contract is used to verify and store the trusted data packet; the responsibility chain contract is used to generate a responsibility chain record when the human-machine collaborative auditing module submits the manual audit evidence and judgment, and the responsibility chain record includes: an operation type identifier, a digital identity identifier of the responsible entity of the corresponding auditor, a hash value or clue identifier of the original record, and multi-party digital signatures.
[0062] Furthermore, the versioned audit trail includes: the target rule version identifier that triggers the generation of the versioned audit trail, the hash value set of the audit data involved, the data hash value set, and the credibility assessment information summary.
[0063] Furthermore, the versioned intelligent audit engine module is further configured as follows: Based on the timestamp in the trusted data packet and the effective time interval of the versioned audit rule, determine the target rule version identifier that should be invoked at the current audit time point; During the automated analysis process, a differentiated verification strategy for the data to be audited is dynamically generated and executed based on the credibility assessment information. The target rule version identifier, the automated audit findings generated after executing the differentiated verification strategy, and the set of hash values of the audit data involved are used together as the versioned audit clues.
[0064] Furthermore, the human-machine collaborative auditing module is further configured as follows: Upon receiving the versioned audit trail, based on the credibility assessment information summary contained in the versioned audit trail, a list of associated differentiated audit tasks is dynamically organized and presented to the auditors. The list of differentiated audit tasks includes the data range to be reviewed, the suggested audit procedure types, and priorities. The auditors perform the differentiated audit tasks through the interactive interface and submit corresponding manual audit evidence and manual audit judgments.
[0065] Furthermore, the audit report and verification service module is further configured as follows: When generating the final audit report, the versioned audit clues, manual audit judgments and corresponding responsibility chain records associated with the complete audit trajectory are structurally linked to generate an embedded evidence index with verifiable pointing relationships.
[0066] Furthermore, the audit report and verification service module is also equipped with a verification interface; The verification interface is configured to: receive a report to be verified submitted by a user, and by parsing the embedded evidence index, retrieve and compare the hash value of the corresponding original record stored in the blockchain evidence storage and responsibility chain module to verify the integrity and tamper-proof nature of the report to be verified.
[0067] Furthermore, the versioned intelligent audit engine module is also configured to: respond to a historical replay instruction, based on a specified historical time point, invoke the versioned audit rules that are in effect at the historical time point and the trusted data snapshot based on the historical time point, recalculate the historical audit conclusions, and generate historical replay audit results for comparison and analysis. The historical replay audit results are stored through the blockchain evidence storage and responsibility chain module and used for comparison and analysis.
[0068] Furthermore, the audit report and verification service module is configured as follows: Based on the credibility assessment information, the certainty level of the versioned audit rules, and the review level of the manual audit judgment, the evidence strength level corresponding to the audit conclusion is calculated, and when generating the final audit report, the evidence strength level is marked and embedded into the evidence index of the report.
[0069] Example 2: This embodiment applies to Embodiment 1, providing an application scenario for the aforementioned intelligent auditing system based on blockchain technology. It assumes a large group enterprise with multiple subsidiaries and decentralized business systems. The core ERP records sales orders and outbound information, the financial system records vouchers and accounts receivable, and the bank provides receipts and transaction APIs. The auditing firm needs to verify the authenticity of revenue recognition, its consistency with cash receipts, and the existence of unusually large transactions or cross-period recognitions during the annual audit. Simultaneously, the audit process must be traceable, accountability must be assignable, and reports must be verifiable.
[0070] 1) Reliable data collection and scoring; The system first connects to the group's ERP, financial system, and bank API to collect audit-related transaction data, such as transaction number, amount, and date. For each piece of data to be audited, the system calculates its hash value and performs a credibility assessment based on multiple dimensions. The generated credibility assessment information, along with the data hash value and timestamp, forms a "trusted data packet" and is sent to the blockchain evidence storage and responsibility chain module. These dimensions include: the authority of the data source, the type of data collection interface, the redundancy of multiple nodes in the data acquisition method, and historical accuracy.
[0071] 2) Blockchain-based evidence storage and responsibility chain recording; The blockchain-based evidence storage and accountability module receives and stores trusted data packets via smart contracts, ensuring the immutability of all audit data. Whenever auditors perform critical operations, such as flagging a transaction as a risk indicator or confirming its legality, the system records the primary and secondary responsible parties for the operation and links this information to the audit data, forming a accountability chain record. All these records are uploaded to the blockchain via smart contracts, ensuring clear attribution of responsibility and the integrity of the audit data. The primary responsible party is the auditor who initiated the audit operation; the secondary responsible party is the auditor who reviewed the operation.
[0072] 3) Version-based rule-driven automated analysis and clue generation; The versioned intelligent audit engine module automatically selects the effective versioned audit rules based on the audit time point and analyzes the collected audit data in conjunction with the data's credibility assessment information. During the analysis, the system dynamically adjusts the verification strategy based on the credibility assessment information. For low-credibility data, the system requires more rigorous review to ensure the accuracy of the analysis. The audit clues generated by the system include rule version identifiers, data hash value sets, and credibility assessment information summaries. All results are recorded and used as the basis for subsequent audit work.
[0073] 4) In-depth auditing and evidence consolidation through human-machine collaboration; The human-machine collaborative auditing module receives audit leads generated by the versioned auditing engine and dynamically organizes and presents a differentiated audit task list based on the credibility assessment information in the leads. The list includes the data scope requiring review, suggested audit procedure types, and task priorities. Auditors perform in-depth audits based on the task list on the interactive interface and submit all evidence and human audit judgments generated during the audit process to the blockchain evidence storage and accountability chain module for storage. Human evidence and judgments generate corresponding accountability chain records through the accountability chain contract, ensuring the traceability of the human audit process.
[0074] 5) Report generation, evidence indexing, and verification; The audit report and verification service module generates a final audit report based on the complete audit trail stored on the blockchain. During report generation, the system structurally correlates relevant versioned audit leads, human audit judgments, and chain-of-responsibility records to form an embedded evidence index. This index, along with the report summary hash value, is published, making the report verifiable. External users can submit reports for verification through a verification interface. The system parses the evidence index in the report, retrieves and compares the hash values of the original records stored on the blockchain, thereby verifying the report's integrity and tamper-proof nature, and tracing the evidence chain upon which the report is based within authorized scope. The complete audit trail includes: trusted data packets, versioned audit leads, human audit evidence and judgments, and chain-of-responsibility records.
[0075] In this application scenario, blockchain technology not only ensures the immutability of data and the audit process, but also guarantees the automation of the audit process and the clear attribution of responsibility through smart contracts. The system's automated analysis and versioning rules drive the generation of audit clues, improving audit efficiency and ensuring the traceability and transparency of audit conclusions, which meets the high requirements of modern enterprise auditing for data credibility, clear responsibility, and verification of audit results.
[0076] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.
[0077] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.
Claims
1. An intelligent auditing system based on blockchain technology, characterized in that, The system includes: The trusted data collection and scoring module is used to connect to data sources and collect data to be audited; calculate the hash value of the data to be audited, and generate the corresponding trusted assessment information based on multiple preset trusted assessment dimensions; and send the trusted data packet including the hash value, timestamp and trusted assessment information of the data to be audited to the blockchain evidence storage and responsibility chain module. The blockchain evidence storage and responsibility chain module is equipped with smart contracts to receive and store the trusted data packets, forming a tamper-proof audit data chain. When performing key audit operations, it records the information of the first responsible party initiating the audit operation and the information of the second responsible party confirming the audit operation, forming a responsibility chain record associated with the audit data. The versioned intelligent audit engine module stores versioned audit rules with version identifiers that are effective within a specified time interval. This module is configured to: obtain the trusted data packet from the blockchain evidence storage and responsibility chain module, call the effective versioned audit rules corresponding to the current audit time point, and automatically analyze the data to be audited in conjunction with the credibility assessment information to generate versioned audit clues. The human-machine collaborative auditing module is used to receive the versioned audit clues, provide auditors with an interactive interface to execute in-depth auditing procedures based on the versioned audit clues, and submit the manual audit evidence and manual audit judgments generated by the auditors in the interactive interface to the blockchain evidence storage and responsibility chain module for evidence storage, so as to update the audit data chain and the responsibility chain records. The audit report and verification service module is used to generate a final audit report based on the complete audit trajectory stored in the blockchain evidence storage and responsibility chain module, which includes the trusted data packet, the versioned audit clues, the manual audit evidence and the manual audit judgment, and the responsibility chain record, and to calculate and publish the summary hash value of the final audit report.
2. The intelligent auditing system based on blockchain technology according to claim 1, characterized in that, The preset multiple credibility assessment dimensions include: the authority level of the data source, the type of data collection interface, the redundancy of multiple nodes in the data acquisition method, and the historical accuracy of the data provided by the data source.
3. The intelligent auditing system based on blockchain technology according to claim 1, characterized in that, The smart contracts deployed in the blockchain evidence storage and responsibility chain module specifically include an evidence storage contract and a responsibility chain contract. The evidence storage contract is used to verify and store the trusted data packet. The responsibility chain contract is used to generate a responsibility chain record when the human-machine collaborative auditing module submits the manual audit evidence and judgment. The responsibility chain record includes: an operation type identifier, a digital identity identifier of the responsible entity of the corresponding auditor, a hash value or clue identifier of the original record, and multi-party digital signatures.
4. The intelligent auditing system based on blockchain technology according to claim 1, characterized in that, The versioned audit trail includes: the target rule version identifier that triggers the generation of the versioned audit trail, the set of hash values of the audit data involved, the set of data hash values, and a summary of credibility assessment information.
5. The intelligent auditing system based on blockchain technology according to claim 4, characterized in that, The versioned intelligent audit engine module is further configured as follows: Based on the timestamp in the trusted data packet and the effective time interval of the versioned audit rule, determine the target rule version identifier that should be invoked at the current audit time point; During the automated analysis process, a differentiated verification strategy for the data to be audited is dynamically generated and executed based on the credibility assessment information. The target rule version identifier, the automated audit findings generated after executing the differentiated verification strategy, and the set of hash values of the audit data involved are used together as the versioned audit clues.
6. The intelligent auditing system based on blockchain technology according to claim 1, characterized in that, The human-machine collaborative auditing module is further configured as follows: Upon receiving the versioned audit trail, based on the credibility assessment information summary contained in the versioned audit trail, a list of associated differentiated audit tasks is dynamically organized and presented to the auditors. The list of differentiated audit tasks includes the data range to be reviewed, the suggested audit procedure types, and priorities. The auditors perform the differentiated audit tasks through the interactive interface and submit corresponding manual audit evidence and manual audit judgments.
7. The intelligent auditing system based on blockchain technology according to claim 1, characterized in that, The audit report and verification service module is configured as follows: When generating the final audit report, the versioned audit clues, manual audit judgments and corresponding responsibility chain records associated with the complete audit trajectory are structurally linked to generate an embedded evidence index with verifiable pointing relationships.
8. The intelligent auditing system based on blockchain technology according to claim 7, characterized in that, The audit report and verification service module is also equipped with a verification interface; The verification interface is configured to: receive a report to be verified submitted by a user, and by parsing the embedded evidence index, retrieve and compare the hash value of the corresponding original record stored in the blockchain evidence storage and responsibility chain module to verify the integrity and tamper-proof nature of the report to be verified.
9. The intelligent auditing system based on blockchain technology according to claim 5, characterized in that, The versioned intelligent audit engine module is also configured to: respond to a historical replay instruction, based on a specified historical time point, invoke the versioned audit rules that are in effect at the historical time point and the trusted data snapshot based on the historical time point, recalculate the historical audit conclusions, and generate historical replay audit results for comparison and analysis. The historical replay audit results are stored through the blockchain evidence storage and responsibility chain module and used for comparison and analysis.
10. The intelligent auditing system based on blockchain technology according to claim 7, characterized in that, The audit report and verification service module is further configured as follows: Based on the credibility assessment information, the certainty level of the versioned audit rules, and the review level of the manual audit judgment, the evidence strength level corresponding to the audit conclusion is calculated, and when generating the final audit report, the evidence strength level is marked and embedded into the evidence index of the report.