Whole life education information management system with credible audit tracking ability

Abstract

The present application relates to the technical field of education informatization and data management, in particular to a whole life cycle education information management system with trusted audit tracking capability, comprising: business access and arrangement, data verification and processing, double-line data recording, state reconstruction and feedback module; each business end education data is converted into a unified event and formed into a queue according to dependence, after verifying identity, authorization, approval, dependence and field integrity, it is written into an audit database and a business state library synchronously, the historical state is reconstructed based on the audit event sequence playback, and the alternative value query data package is output according to the query role, the present application realizes that the education data is traceable, auditable and trusted.

Description

Technical Field

[0001] This invention relates to the field of educational informatization and data management technology, specifically to a full life-cycle education information management system with reliable audit traceability capabilities. Background Technology

[0002] In the existing process of talent cultivation information management, the following technical deficiencies exist when processing cross-departmental and cross-system business data.

[0003] Different business systems use their own independent data structures with inconsistent field standards and unstable event arrival order, which can easily lead to situations where subsequent records enter the system before the preceding business is completed. Existing systems mostly use a direct overwrite update mode to save business results, which makes it difficult to completely retain historical change trajectories when revisions, additions, or cancellations occur, and it is impossible to accurately reconstruct the business status at any historical point in time based on a unified time sequence. In the query feedback process, existing solutions have problems with coarse field permission control, insufficient desensitization and proof capabilities, and it is difficult to balance data availability, auditability, and minimum disclosure requirements. Summary of the Invention

[0004] The purpose of this invention is to provide a full lifecycle education information management system with reliable audit trail capabilities, avoiding direct data overwriting that leads to the inability to trace historical states, solving the problem of temporal sequence disorder in cross-system data flow, and achieving the coexistence of anonymized display and evidence verification. This invention can be achieved through the following technical solutions:

[0005] The business access and orchestration module is used to receive educational data submitted by various business terminals and convert it into unified events containing object identifiers. The unified events are grouped and sorted according to the object identifiers, and a queue of events to be processed is formed after the prerequisite dependencies are satisfied.

[0006] The data verification and processing module is used to perform compliance verification on events in the event queue to be processed according to preset rules, and assign a valid flag or a flag to be reviewed based on the verification results.

[0007] The dual-line data recording module is used to write events carrying valid flags into the audit database and the business status database in the same recording transaction to generate event feature values, obtain the direct predecessor events corresponding to the preceding dependencies of the event, associate the event feature values ​​with the feature values ​​of the direct predecessor events and write them into the audit database, and record the digital signature and version reference; the business status database is used to store the status index.

[0008] The State Reconstruction and Feedback module is used to respond to historical state query commands carrying query roles. Based on the object identifier, it retrieves the event sequence from the audit database, performs state replay to obtain the reconstructed business state, determines the invisible fields in the reconstructed business state according to the permissions corresponding to the query role, generates corresponding replacement values ​​for the invisible fields, and combines the output into a query data packet.

[0009] Optionally, the business access and orchestration module is specifically used to receive the raw data from each business end and convert it into a unified event containing object identifier, operation type, business validity time, submission time, operator identity, approval node information and prerequisite dependency identifier;

[0010] Associate the operator's identity with their role, authorization period, and purpose identifier, and verify their processing permissions;

[0011] Grouping and sorting according to the rule of priority based on the effective time of business, followed by the submission time;

[0012] When there is an event whose prerequisite dependency is not satisfied, the event is suspended until the dependency is satisfied before being enqueued to output the queue of events to be processed.

[0013] Optionally, the compliance verification of the data verification and processing module specifically includes:

[0014] The verification results for identity and role matching, authorization period validity, approval node integrity, prerequisite dependency satisfaction, data field integrity, and business content consistency are determined respectively.

[0015] When all the above verification results meet the corresponding rules, a valid flag is assigned.

[0016] Optionally, when a pending review flag is assigned, the data verification and processing module is specifically used to return the event and request re-authorization or additional approval if the failure is due to mismatch between identity and role, expiration of authorization period, or non-compliance of purpose identifier.

[0017] If the approval fails due to missing approval nodes, unmet prerequisites, or missing prerequisite data, the event will be suspended and the approval completion, prerequisite data completion, or dependency confirmation will be triggered.

[0018] If the failure is due to missing data fields, incorrect field format, or inconsistencies in business content, it will be marked as pending manual review and a review task will be generated.

[0019] The completed review conclusion will be treated as a new event and re-verified and subsequently recorded.

[0020] Optionally, the business status update rules adopted by the dual-line data recording module include:

[0021] When the operation type is "add", the new data will be added to the valid collection;

[0022] When the operation type is revision, the old version data is retained and the new version data is generated. The old version data is given an invalidation flag in the business status database and its history is retained in the audit database.

[0023] When the operation type is supplementary recording, the supplementary recording time and the business validity time are recorded, and the corresponding status index is updated accordingly.

[0024] Optionally, the dual-line data recording module is also used to serialize the object identifier, operation type, content field and time field of the event to obtain a standardized code;

[0025] Sort and concatenate the feature values ​​of direct predecessor events according to submission time and unified event identifier;

[0026] Feature values ​​are calculated based on the results of normalized encoding and sorting concatenation to generate event feature values.

[0027] Optionally, when generating the query data package, the state reconstruction and feedback module is specifically used to determine the set of visible fields and the set of invisible fields in the reconstructed business state based on the field permission table corresponding to the query role, and generate field permission identifiers.

[0028] For invisible fields, output the corresponding desensitized value, range value, category value, or hash value as a replacement value according to the field type;

[0029] The event characteristic value, digital signature, operator permission information, field permission identifier, and version reference are combined to generate a query data package.

[0030] Optionally, the data verification and processing module also includes a parameter correction unit, used to receive historical verification statistics including the number of times each verification item passes, the number of times it fails, and the number of times it passes the re-verification;

[0031] Based on historical verification statistics, the judgment weights of each verification item and the comprehensive risk classification threshold are dynamically adjusted according to preset rules to obtain updated judgment parameters.

[0032] Write the updated decision parameters into the parameter configuration table for later use.

[0033] Compared with the prior art, the present invention has the following advantages:

[0034] 1. This invention converts heterogeneous educational data submitted by various business terminals into standardized events and performs causal orchestration by grouping and sorting by object identifiers and enqueuing data after satisfying prerequisite dependencies. This effectively solves the problems of inconsistent data fields from multiple sources, unstable arrival order, and subsequent records entering the system before the prerequisite business is completed.

[0035] 2. This invention utilizes a multi-dimensional verification mechanism, including identity and role matching, authorization period, approval node integrity, prerequisite dependency satisfaction, and field integrity, to accurately identify and triage abnormal events, reducing business interference caused by false interception and invalid transfer;

[0036] 3. This invention adopts a dual-line data recording method of audit database and business status database, and completes synchronous writing and generates event feature values ​​and version references in the same record transaction, so that the historical change trajectory can be completely preserved, avoiding the loss of historical evidence and the problem of untraceable status caused by direct overwrite updates. Attached Figure Description

[0037] The above and other objects, features and advantages of the present invention will become clearer from the following description of embodiments of the invention with reference to the accompanying drawings, in which:

[0038] Figure 1 This is a schematic diagram of the modules of the full life-cycle education information management system with trusted audit trail capabilities provided in the embodiments of this application. Detailed Implementation

[0039] To make the above features and advantages of the present invention more apparent and understandable, exemplary embodiments are described below in detail with reference to the accompanying drawings.

[0040] Please see Figure 1 A full-lifecycle education information management system with reliable audit trail capabilities, including:

[0041] The business access and orchestration module is used to receive educational data submitted by various business terminals and convert it into unified events containing object identifiers. The unified events are grouped and sorted according to the object identifiers, and a queue of events to be processed is formed after the prerequisite dependencies are satisfied.

[0042] The data verification and processing module is used to perform compliance verification on events in the event queue to be processed according to preset rules, and assign a valid flag or a flag to be reviewed based on the verification results.

[0043] The dual-line data recording module is used to write events carrying valid flags into the audit database and the business status database in the same recording transaction to generate event feature values, obtain the direct predecessor events corresponding to the preceding dependencies of the event, associate the event feature values ​​with the feature values ​​of the direct predecessor events and write them into the audit database, and record the digital signature and version reference; the business status database is used to store the status index.

[0044] The State Reconstruction and Feedback module is used to respond to historical state query commands carrying query roles. Based on the object identifier, it retrieves the event sequence from the audit database, performs state replay to obtain the reconstructed business state, determines the invisible fields in the reconstructed business state according to the permissions corresponding to the query role, generates corresponding replacement values ​​for the invisible fields, and combines the output into a query data packet.

[0045] The system acquires educational data submitted by various business units, and according to a unified transformation rule, specifically by parsing the data dictionary, mapping and aligning the fields input from heterogeneous systems to a standard attribute set, filling in default values ​​for missing non-mandatory fields, formatting and transforming the educational data, generating a unified event, and adding it to the pending event queue.

[0046] The events in the queue are subjected to multi-dimensional verification according to preset rules to determine the verification results; the events that pass the verification are synchronously written to the audit database and the business status database to generate event feature values ​​and status indexes.

[0047] In response to the query request, the reconstructed business status is determined based on the event sequence replaying in the audit database. The specific process is as follows: with the initial empty state as the baseline, the events are processed sequentially according to the order of their valid business times. When the valid business times are the same, they are processed according to the order of their submission times. If the two are still the same, they are processed according to the fixed lexicographical order of the unified event identifier to ensure that the replay order is unique.

[0048] The operation type of each event in the sequence is applied sequentially to the baseline to overlay the state. If a new event is added, a record is added. If a revision is encountered, the old field is overwritten with the corresponding field of the new version. If a supplementary event is encountered, the corresponding record is inserted or corrected to the corresponding historical time point according to its business validity time, and the state cascading update control flow is triggered. Specifically, the affected event subset is traversed in a forward direction from the supplementary event time point to the current time point. The new state output by the supplementary event is used as the boundary condition to re-execute the state overlay operation to update the subsequent state index.

[0049] If a logical conflict occurs during the traversal calculation, causing the baseline state to overflow abnormally or the reference to become invalid, the engineering degradation rule is triggered. The reconstruction result of the current object is tagged with time-series dirty data and the cascading replay operation is terminated. Instead, the stable and valid version that was successfully written to the database and passed the compliance verification before the conflict is returned, until the calculation reaches the target query point, and the query data packet is output according to the permissions.

[0050] Among them, a unified event refers to a standardized data structure that includes key elements such as object identifier, operation type, and business validity time, such as a new scholarship issuance record for student ID 2023001; and writing to the audit database and business status database in the same record transaction means using a single atomic commit as the boundary for external effectiveness: the current event is only visible to the query side when both the audit database write and the business status database write are successfully completed.

[0051] If any write operation fails, the current record operation will be rolled back and kept invisible to both databases, thereby avoiding timing conflicts such as the audit track being updated but the business projection track not being updated, or vice versa.

[0052] This invention preserves an immutable event stream through a dual-line data recording mechanism, transforming historical changes into audit evidence that can be traced back at any time. This embodiment can achieve the unity of business reproducibility and audit non-repudiation based on the dual-line event recording model and historical state reconstruction mechanism, avoiding the inability to trace historical states due to direct data overwriting.

[0053] In this embodiment, the business access and orchestration module is specifically used to receive the raw data from each business terminal and convert it into a unified event containing object identifier, operation type, business validity time, submission time, operator identity, approval node information and prerequisite dependency identifier.

[0054] Associate the operator's identity with their role, authorization period, and purpose identifier, and verify their processing permissions;

[0055] Grouping and sorting according to the rule of priority based on the effective time of business, followed by the submission time;

[0056] When there is an event whose prerequisite dependency is not satisfied, the event is suspended until the dependency is satisfied before being enqueued to output the queue of events to be processed.

[0057] The system acquires raw data from academic affairs, student affairs, and psychological support units, transforms the raw data according to standardized specifications, and generates unified events. Among these, operator permission information refers to the permission context bound by the operator's identity, role, authorization period, and purpose identifier, such as the role of a counselor, the authorization period ending at the end of the semester, and the purpose limited to scholarship and grant evaluation.

[0058] This invention captures the dependencies between events through an event causal orchestration mechanism, temporarily suspending events that are not met due to unmet preconditions, such as suspending the evaluation event before the results are released, and transforming the disordered data stream into a queue of pending processing that follows the business sequence.

[0059] Furthermore, the conversion is performed according to a fixed field mapping table during the specific implementation: first, the source system identifier is read and the corresponding mapping template is located; then, the source field names are mapped one by one to the unified field names; for enumerated fields, values ​​are replaced according to a preset code table; for missing non-mandatory fields, default values ​​are written; and for unrecognizable redundant fields, they are written to the extended field area and do not participate in this round of verification.

[0060] The determination of the satisfaction of the preceding dependency is carried out by an explicit table lookup method in the specific implementation: for each preceding dependency identifier in the event, the audit database is queried to see if there is a predecessor event with a matching object identifier and a valid flag of true. If all of them exist, it is recorded as a dependency satisfaction. Otherwise, the current event is written to the pending table, and the set of missing dependency identifiers is recorded in the pending table.

[0061] The suspend table is configured with dynamic wake-up control logic based on the publish-subscribe model: when any subsequent event is verified to obtain a valid flag and is successfully stored in the database, its unified event identifier is broadcast to the subscription channel as a trigger event; after receiving the broadcast, the suspend control flow concurrently retrieves the missing dependency set containing this flag in the suspend table and performs a set removal operation. Once the number of elements in the missing dependency set of a suspended event drops to zero, the state machine of the event is driven to flip from dependency suspension to ready, and it is pushed back into the queue of pending events.

[0062] The sorting key of the pending event queue is fixed as follows: priority is given to the business validity time under the object identifier group, followed by the submission time, and then the unified event identifier. The event sequence entering the audit database follows the same stable order. This embodiment of the invention can realize the orderly flow of events based on causal orchestration, avoiding business logic conflicts caused by disordered timing.

[0063] In this embodiment, the compliance verification of the data verification and processing module specifically includes:

[0064] The verification results for identity and role matching, authorization period validity, approval node integrity, prerequisite dependency satisfaction, data field integrity, and business content consistency are determined respectively.

[0065] When all the above verification results meet the corresponding rules, a valid flag is assigned.

[0066] Obtain the unified event to be processed, perform structured verification on the event according to the preset data compliance verification rules, and determine the final verification result;

[0067] In practical implementation, a credibility score can be further calculated to characterize the overall risk level of the unified event with index number i currently being processed. The score is obtained by weighted summation of multiple compliance items, using the following formula:

[0068]

[0069] Among them, identity compliance item Used to assess the match between the operator's identity and the permissions required for the current business, and to approve complete items. Used to assess the ratio of the number of approved nodes to the number of nodes that should be met, causal completeness. Used to evaluate the satisfaction of preceding dependent events, data completeness item Used to assess the degree of match between submitted data fields and required fields;

[0070] For causal integrity verification involving rigid physical constraints, the result does not participate in the linear weighting; instead, it acts as a gating switch for the scoring operator, activating the credibility score only if and only if the preceding dependencies are fully satisfied. The computing engine;

[0071] At the same time, weight parameters are introduced. , , , This is used to dynamically balance the contribution ratio of each verification dimension in the comprehensive evaluation. An example value of 0.25 is used, and each weight is limited to between 0 and 1, with the sum equal to 1, to ensure the reliability score. It remains within the 0 to 1 range;

[0072] Specifically, the aforementioned identity compliance items Complete approval item , causal complete item Data completeness item In practical implementation, it can be mapped to a quantitative score between 0 and 1 according to preset rules: for example, for complete approval items If the number of approval nodes that have been processed is greater than or equal to the number of approval nodes that should be satisfied, then a value is assigned. If it is less than , the value is the ratio of the number of nodes that have been transferred to the number of nodes that should be satisfied;

[0073] If the required number of nodes is 0, then the complete item should be approved. We directly take 1 to avoid the denominator of the ratio being zero; for complete causal terms... If all direct preceding dependent events have been assigned a valid flag, the value is 1; otherwise, it is 0.

[0074] Furthermore, to ensure a one-to-one correspondence with the rigid verification conditions, the system outputs Boolean results for identity and permission compliance, approval node completeness, prerequisite dependency satisfaction, data field completeness, and consistency between business content and the approved business scope. , , , , Among them, the identity and permission compliance is rigidly determined through multi-level conditional branches: the first branch checks whether the set of operator roles contains the allowed roles corresponding to the event. If not, the process is interrupted and determined to be false.

[0075] The secondary branch retrieves the timestamp of the event recorded by the business side. Read the fixed phase deviation between the standard server time and the business time recorded during the system synchronization handshake. It also reads the pre-calibrated clock drift rate based on crystal oscillator offset calculation from the device hardware ledger via the underlying interface. The alignment time is calculated using a clock drift compensation operator. :

[0076]

[0077] Determine alignment time Whether it strictly falls within the closed interval of the authorized valid time axis; if it exceeds the boundary, it is judged as false; the last-level branch checks whether the business purpose identifier matches the predefined allowed purpose set of the event type;

[0078] The output is true only if no exception interception is triggered in any decision branch; the completeness of the approval node is determined by comparing the list of required approval nodes corresponding to the event type with the list of submitted approval nodes. If all required nodes appear and the sequence number is not reversed, the output is true; the satisfaction of the prerequisite dependency is determined by querying the valid flags of the direct predecessor events item by item; the completeness of the data field is determined by checking the list of required fields, the list of field formats, and the list of value ranges for the event type.

[0079] The consistency between the business content and the approved business scope is determined by the following fixed rules: the approved business scope table is read from the most recent approval event. The approved business scope table includes at least one or more of the following: allowed object identifier, allowed business category, allowed time interval, and allowed amount interval or allowed status set.

[0080] The object identifier, business category, business validity period, and key business fields of the current event are compared item by item with the approval business scope table. Only when all restricted fields fall within the corresponding allowed range will the approval be granted. Only accept the truth; otherwise, accept the false.

[0081] For example, in a revision event for award / grant disbursement, if the current revised amount exceeds the maximum approved amount recorded in the most recent approved event, or if the effective period of the business exceeds the approved time range, then it is directly determined that the business content is inconsistent with the scope of the approved business; the final verification pass condition adopts rigid logic:

[0082]

[0083] Credibility scores are used for review priority ranking, risk warning, or statistical analysis. They do not replace the rigid judgment logic that a pass result is obtained only when all verification results meet the corresponding rules. When any required verification item fails to meet the corresponding rule, the current event should not be directly judged as pass, but should enter the process of suspension, rejection, or waiting for manual review.

[0084] This invention uses fine-grained, multi-dimensional structured verification to locate verification steps that do not meet requirements; it can achieve accurate data verification based on a combination of multi-dimensional verification and risk quantification.

[0085] In this embodiment, when a pending review flag is assigned, the data verification and processing module is specifically used to return the event and request re-authorization or additional approval if the failure is caused by mismatch between identity and role, expiration of authorization period, or non-compliance of purpose identifier.

[0086] If the approval fails due to missing approval nodes, unmet prerequisites, or missing prerequisite data, the event will be suspended and the approval completion, prerequisite data completion, or dependency confirmation will be triggered.

[0087] If the failure is due to missing data fields, incorrect field format, or inconsistencies in business content, it will be marked as pending manual review and a review task will be generated.

[0088] The completed review conclusion will be treated as a new event and re-verified and subsequently recorded.

[0089] The system retrieves abnormal events marked as pending review and, based on the different types of abnormalities, categorizes and processes these events, generating corresponding review tasks or completion requests. The abnormality handling mechanism refers to the corresponding remedial actions triggered by different errors, such as triggering approval completion if an approval node is missing, or transferring the process to manual review if a field is missing.

[0090] This invention transforms the offline error correction process into an online traceable event stream by resending the review conclusion after processing as a new event to the verification module. This embodiment can realize the engineering applicability of data error correction based on the abnormal closed-loop processing mechanism, avoiding business stagnation caused by a single interception.

[0091] In this embodiment, the data verification and processing module further includes a parameter correction unit, which is used to receive historical verification statistics including the number of times each verification item passes, the number of times it fails, and the number of times it passes the re-verification.

[0092] Based on historical verification statistics, the judgment weights of each verification item and the comprehensive risk classification threshold are dynamically adjusted according to preset rules to obtain updated judgment parameters.

[0093] Write the updated decision parameters into the parameter configuration table for later use.

[0094] Obtain historical verification statistics from historical review results, and adjust the judgment weights of each verification item according to the preset judgment weight adjustment rules to determine the updated judgment parameters.

[0095] In the specific implementation, the number of times each verification item passes, fails, and passes upon re-review within a preset statistical period is counted, and the corresponding results are normalized to the same order of magnitude between 0 and 1 to obtain the historical verification contribution value characterizing the interception effectiveness of a single item. A higher contribution value indicates that the verification item is more effective in identifying abnormal events in historical samples. To ensure the executability of the calculation process, in the specific implementation, each verification item... Follow these steps:

[0096] Read the number of times this verification item passes within the statistical period. Number of times failed and the number of times the review was approved Here, the number of times the review was approved will be recorded. The statistical scope is limited to the number of events that were initially blocked due to failing the verification item, but were later manually reviewed and determined to be eligible for continued processing. Therefore, it must satisfy the following condition. If the statistical value is greater than the number of failures when summarizing across tables... For abnormal records, then according to Cut off the portion before calculating;

[0097] Number of passes Number of times failed Add them together to get the total number of samples. If the total number of samples If the value is 0, then the parameters of the previous cycle remain unchanged;

[0098] Calculate the interception rate and the review and correction rate When the number of times it fails When the value is 0, the review and correction rate will be... Recorded as 0;

[0099] Using the aforementioned interception rate and review and correction rate, calculate the historical verification contribution value of this verification item. This historical verification contribution value The larger the value, the lower the proportion of exceptions intercepted by the verification item that are overturned upon review;

[0100] Historical verification contribution value for all verification items The total contribution value is obtained by summing. If the total contribution value If greater than 0, then follow Normalization is performed to obtain the updated judgment weights, if the total contribution value If the value is 0, the original weight remains unchanged;

[0101] Based on this contribution value, the judgment weight of the verification item with a pass rate less than the first preset threshold and a review and correction rate greater than the second preset threshold is reduced by the first preset step size, and the judgment weight of the verification item with a pass rate less than the first preset threshold and a review and correction rate less than the second preset threshold is increased by the second preset step size.

[0102] Comprehensive risk grading threshold used to divide the review priority queue Based on the proportion of abnormal events in the current statistical period and preset business risk tolerance coefficient Perform segmented corrections; among which, The range of values ​​read from the system configuration table is within A fixed scalar; for example, when At that time, threshold Set to 0.80, when At that time, the threshold The value was increased to 0.85. At that time, the threshold Increased to 0.90; the threshold range is strictly limited to between 0 and 1;

[0103] During updates, a fixed table writing process is followed: a parameter record containing the statistical period number, new weights for each verification item, new thresholds, and effective time is generated; the parameter record is written to the parameter configuration table corresponding to the data compliance verification rules; for subsequent events that enter the queue for review but have not yet been assigned to a human handler, the latest effective parameter record is read and prioritized.

[0104] It should be noted that the updated judgment parameters, including the judgment weights of each verification item and the comprehensive risk classification threshold, are only used to dynamically adjust the timing priority of a unified event in the manual intervention review queue or trigger a high-risk warning based on the comparison results of its credibility score and the comprehensive risk classification threshold when the unified event fails to pass due to rigid rules. They do not change the passing conditions of rigid rules such as identity permissions, approval nodes, prerequisite dependencies and field integrity at all.

[0105] This invention introduces feedback from historical statistical data to adaptively adjust and dynamically adapt to fluctuations in the quality of business data at different times; it can achieve adaptive optimization of verification parameters based on historical feedback, ultimately solving the problem of uneven allocation of verification resources and high false interception rate caused by the lack of dynamic adjustment mechanism in the prior art.

[0106] In this embodiment, the business status update rules adopted by the dual-line data recording module include:

[0107] When the operation type is "add", the new data will be added to the valid collection;

[0108] When the operation type is revision, the old version data is retained and the new version data is generated. The old version data is given an invalidation flag in the business status database and its history is retained in the audit database.

[0109] When the operation type is supplementary recording, the supplementary recording time and the business validity time are recorded, and the corresponding status index is updated accordingly.

[0110] The dual-line data recording module is also used to serialize the object identifier, operation type, content field, and time field of an event to obtain a standardized code;

[0111] Sort and concatenate the feature values ​​of direct predecessor events according to submission time and unified event identifier;

[0112] Feature values ​​are calculated based on the results of normalized encoding and sorting concatenation to generate event feature values;

[0113] When generating the query data package, the state reconstruction and feedback module is specifically used to determine the set of visible and invisible fields in the reconstructed business state based on the field permission table corresponding to the query role, and generate field permission identifiers.

[0114] For invisible fields, output the corresponding desensitized value, range value, category value, or hash value as a replacement value according to the field type;

[0115] The event characteristic value, digital signature, operator permission information, field permission identifier, and version reference are combined to generate a query data package.

[0116] Acquire unified events carrying valid flags, classify and record events according to business status update rules, and generate updated status indexes; among them, dual-line data recording refers to maintaining synchronous updates of the audit track and business projection track in the same transaction. For example, for revision operations, the old version is marked as invalid in the business status database but retains a complete record in the audit database.

[0117] Furthermore, the completion determination of synchronous updates adopts a consistent submission threshold: the event status is switched from processing to effective only when the audit track is successfully written, the business projection track is successfully written, and the corresponding version reference is established; if any step fails, the writing result of this round is revoked and the previous stable version is retained, thereby ensuring that the query end will not observe a half-completed state.

[0118] This invention transforms single-dimensional state maintenance into event-stream-based projection calculation by separating the current business state from historical fact records. This embodiment can permanently retain historical evidence based on state projection rules, avoiding legitimate modification operations from compromising audit integrity.

[0119] The core fields of a unified event are obtained. Based on normalized serialization and hash operations, features are extracted from the event content and its predecessor relationships to generate event feature values. In the specific implementation, feature values ​​used to solidify the causal dependencies of events are calculated.

[0120]

[0121] Regarding the unified events currently being processed Generate normalized codes to eliminate data format differences during the cross-system flow of this event. ;

[0122] Specifically, the key names of the content field and time field of the current event are arranged in ascending alphabetical order. All numeric types are converted to fixed 64-bit hexadecimal strings according to the IEEE 754 standard. Before the conversion, an absolute value zero vector normalization operation is introduced to map -0.0 to +0.0. All time fields are uniformly converted to Coordinated Universal Time timestamp integers in millisecond units and concatenated to output a standardized string.

[0123] Get the currently processing unified event eigenvalues ​​of direct precursor events It generates a sorted and concatenated result that sorts and concatenates the direct predecessor events according to a fixed order of submission time and uniform event identifier. ;

[0124] Extract the total number of events in the set of direct predecessor events as the bounding parameter. Through data splicing operations Normalized coding Vector dimension Combined with sorted results By combining the concatenated data, a hash function H, used to generate an irreversible digest, is applied to calculate the characteristic value of the current event. ;

[0125] Furthermore, when the set of direct predecessor events is empty, the sorted concatenation result... Take a preset empty predecessor marker string so that the first event also has a definite and repeatable characteristic value; when there are multiple direct predecessor events with the same submission time, use a unified event identifier as a stable order key to avoid different summaries from the same predecessor set due to different traversal order.

[0126] This invention characterizes the chain-like causal relationship between events by fusing the feature values ​​of preceding events into the current event summary. This embodiment can realize anti-tampering binding of cross-system data based on chain feature values, ultimately solving the problem in the prior art that the lack of a dependency solidification mechanism makes it difficult to verify the consistency of historical records.

[0127] Obtain the reconstructed business status and query role. Based on the field permission table corresponding to the role, filter and anonymize the fields in the business status to generate a query data package. The query data package refers to a comprehensive feedback carrier containing anonymized business data and cryptographic proof, such as the hash value of invisible fields, the operator's permission information corresponding to the replayed event, the operator's digital signature on the digest, and the snapshot version reference number 1024 of the event in the business status database.

[0128] This invention provides the querying party with the minimum information required to verify the existence of facts through a minimum disclosure proof mechanism, replacing plaintext fields with verifiable cryptographic credentials or restricted representations; this embodiment can achieve the coexistence of de-identified display and evidence verification based on a combination of permission filtering and proof.

[0129] Taking the student registration change and psychological support linkage scenario in the university's full life cycle education information management system as an example; when the academic affairs system generates a downgrade change event, the new data flows into the business access and orchestration module to generate a unified event;

[0130] The system recognizes that the event depends on the grade review event. If the review is not completed, it will be automatically suspended. After the dependency is satisfied and compliance verification is passed, the dual-line data recording module will write the downgrade record to the audit database and project the new valid student status in the business status database. At the same time, it will generate a chain of feature values ​​containing the hash of the preceding grade event.

[0131] When a counselor initiates a query later, the status reconstruction and feedback module replays the data to the current point in time based on the audit database and outputs the query data package according to the counselor's permissions.

[0132] The data packet hides the specific violation details fields and only outputs the category value, but provides a digital signature and version reference; this system implements the processing flow from new data inflow, dependency orchestration, dual-track accounting to compliance de-identification feedback.

[0133] The above are merely preferred embodiments of the present invention. It should be noted that those skilled in the art can make several improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A full-lifecycle education information management system with reliable audit trail capabilities, characterized in that: include: The business access and orchestration module is used to receive educational data submitted by various business terminals and convert it into unified events containing object identifiers, group and sort the unified events according to the object identifiers, and form a queue of events to be processed after satisfying the prerequisite dependencies. The data verification and processing module is used to perform compliance verification on the events in the event queue to be processed according to preset rules, and assign a valid flag or a flag to be reviewed according to the verification result. The dual-line data recording module is used to write events carrying the valid flag into the audit database and the business status database respectively in the same recording transaction to generate event feature values, obtain the direct predecessor events corresponding to the preceding dependencies of the event, associate the event feature values ​​with the feature values ​​of the direct predecessor events and write them into the audit database, and record digital signatures and version references; the business status database is used to store status indexes; The state reconstruction and feedback module is used to respond to historical state query commands carrying query roles, retrieve event sequences from the audit database based on the object identifier, perform state replay to obtain the reconstructed business state, determine the invisible fields in the reconstructed business state according to the permissions corresponding to the query role, generate corresponding replacement values ​​for the invisible fields, and combine them to output a query data packet.

2. The full lifecycle education information management system with reliable audit trail capabilities as described in claim 1, characterized in that, The business access and orchestration module is specifically used to receive raw data from each business terminal and convert it into a unified event containing object identifier, operation type, business validity time, submission time, operator identity, approval node information and prerequisite dependency identifier. Associate the operator's identity with their role, authorization period, and purpose identifier, and verify their processing permissions; The grouping and sorting are based on the rule of prioritizing the effective time of the business and then the submission time. When there is an event whose prerequisite dependency is not satisfied, the event is suspended until the dependency is satisfied before being enqueued, so as to output the queue of events to be processed.

3. The full lifecycle education information management system with trusted audit trail capabilities as described in claim 1, characterized in that, The compliance verification of the data verification and processing module specifically includes: The verification results for identity and role matching, authorization period validity, approval node integrity, prerequisite dependency satisfaction, data field integrity, and business content consistency are determined respectively. When all the above verification results meet the corresponding rules, the valid flag is assigned.

4. The full life-cycle education information management system with reliable audit trail capabilities as described in claim 1, characterized in that, When the pending review flag is assigned, the data verification and processing module is specifically used to return the event and request re-authorization or additional approval if the failure is caused by mismatch between identity and role, expiration of authorization period, or non-compliance of purpose identifier. If the approval fails due to missing approval nodes, unmet prerequisites, or missing prerequisite data, the event will be suspended and the approval completion, prerequisite data completion, or dependency confirmation will be triggered. If the failure is due to missing data fields, incorrect field format, or inconsistencies in business content, it will be marked as pending manual review and a review task will be generated. The completed review conclusion will be treated as a new event and re-verified and subsequently recorded.

5. The full lifecycle education information management system with reliable audit trail capabilities as described in claim 1, characterized in that, The business status update rules adopted by the dual-line data recording module include: When the operation type is "add", the new data will be added to the valid collection; When the operation type is revision, the old version data is retained and the new version data is generated. The old version data is given an invalidation flag in the business status database and its historical record is retained in the audit database. When the operation type is supplementary recording, the supplementary recording time and the business validity time are recorded, and the corresponding status index is updated accordingly.

6. The full lifecycle education information management system with trusted audit trail capabilities as described in claim 1, characterized in that, The dual-line data recording module is also used to serialize the object identifier, operation type, content field and time field of the event to obtain a standardized code; Sort and concatenate the feature values ​​of direct predecessor events according to submission time and unified event identifier; Feature values ​​are calculated based on the results of the standardized encoding and sorting concatenation to generate the event feature values.

7. The full lifecycle education information management system with reliable audit trail capabilities according to claim 1, characterized in that, When generating a query data packet, the state reconstruction and feedback module is specifically used to determine the set of visible fields and the set of invisible fields in the reconstructed business state based on the field permission table corresponding to the query role, and generate field permission identifiers. For the invisible field, output the corresponding desensitized value, interval value, category value, or hash value as the replacement value according to the field type; The query data package is generated by combining the event feature value, the digital signature, the operator permission information, the field permission identifier, and the version reference.

8. The full life-cycle education information management system with trusted audit trail capabilities as described in claim 1, characterized in that, The data verification and processing module also includes a parameter correction unit, which is used to receive historical verification statistics including the number of times each verification item passes, the number of times it fails, and the number of times it passes the re-verification. Based on the historical verification statistics, the judgment weights and comprehensive risk classification thresholds of each verification item are dynamically adjusted according to preset rules to obtain updated judgment parameters. The updated determination parameters are written into the parameter configuration table for later use.

Images