Search

CN-121979836-A - File informatization management method based on AI and dynamic rules and electronic equipment

CN121979836ACN 121979836 ACN121979836 ACN 121979836ACN-121979836-A

Abstract

The invention discloses an AI and dynamic rule-based archive informatization management method and electronic equipment, wherein the method comprises the steps of responding to configuration operation of an archive manager, configuring a user group, and packaging workflow, authority approval workflow and archive utilization workflow based on dynamic rule library configuration information; the method comprises the steps of acquiring metadata of a target file, acquiring metadata of the target file, generating a submitted information packet by adopting a multi-mode AI model, driving an information encapsulation workflow according to the metadata, inquiring a dynamic rule base, converting the submitted information packet into a file information packet based on metadata matching and executing corresponding encapsulation and conversion rules, storing the file information packet into the archive, triggering the file utilization workflow in response to a utilization request of a user on the target file, executing permission verification according to a user group to which the user belongs and an authorized item thereof, and executing retrieval and calling on the archive according to a permission verification result. The invention realizes the dynamic management of archives input, authority management and utilization links through the AI model and the dynamic rules, and has strong universality and high safety.

Inventors

  • HAN ZHIHONG

Assignees

  • 东华软件股份公司

Dates

Publication Date
20260505
Application Date
20251209

Claims (10)

  1. 1. An archive informatization management method based on AI and dynamic rules is characterized by comprising the following steps: Responding to the configuration operation of a archive manager, configuring a user group containing different authorization items, and packaging a workflow, a permission approval workflow and an archive utilization workflow based on predefined dynamic rule base configuration information; Extracting metadata of the newly entered file by adopting a pre-trained multi-mode AI model to generate a submitted information packet; driving the information packaging workflow according to the metadata, wherein the information packaging workflow queries the dynamic rule base, matches and executes corresponding packaging and conversion rules based on the metadata, converts the submitted information packet into a file information packet, and stores the file information packet into an archive; And responding to the utilization request of the user for the target archive, triggering the archive utilization workflow, executing authority verification by the archive utilization workflow according to the user group to which the user initiating the request belongs and the authorized item thereof, and executing retrieval and calling on the archive according to the authority verification result.
  2. 2. The AI-and-dynamic rule-based archive informatization management method of claim 1, wherein said triggering the archive utilization workflow in response to a user's request for utilization of a target archive, said archive utilization workflow performing rights verification based on the user group to which the requesting user belongs and its authorized items, and performing retrieval and invocation of the archive based on the results of the rights verification, comprises: Constructing a dynamic context portrait of the user based on the user group to which the user belongs and the authorization item thereof, wherein the dynamic context portrait at least comprises the semantic intention of the current utilization request, the historical utilization behavior portrait of the user and the risk level of the user group to which the user belongs; triggering a dynamic desensitization engine based on the authority verification result and the dynamic context portrait, and executing file calling after performing self-adaptive desensitization processing on the retrieved target file content; Wherein the desensitization level of the adaptive desensitization process is inversely related to the risk level of the dynamic context portrait.
  3. 3. The AI-and dynamic rule-based archive informatization management method of claim 2, wherein after the archive is retrieved using a workflow, the method further comprises: Triggering an association rule engine to acquire archive association rules in the dynamic rule base in response to completion of the retrieval node, wherein the archive association rules realize association among different archives based on metadata fields; Acquiring an association archive in the archive based on the target archive and the archive association rule; And before the association file is returned to the user, performing secondary authority verification according to the user group to which the user initiating the request belongs and the authorization item thereof.
  4. 4. The AI-and dynamic rule-based archive informatization management method of claim 1, wherein configuring the user group including the different authorization items in response to the configuration operation of the archive manager, and packaging the workflow, the authority approval workflow, and the archive utilization workflow based on the predefined dynamic rule base configuration information, comprises: Responding to the rule configuration operation of the archive manager, updating rules in the dynamic rule base, and respectively configuring rule decision points for the information encapsulation workflow, the authority approval workflow and the archive utilization workflow; And responding to the authority configuration operation of the archive manager, and selecting at least one field in metadata and business index data as an authorization item to establish the user group.
  5. 5. The AI-and dynamic rule-based archive informatization management method of claim 1, wherein driving the information package workflow according to the metadata, the information package workflow querying the dynamic rule base, matching and executing corresponding package and conversion rules based on the metadata, converting the submitted information packet into an archive information packet, and storing the archive information packet in an archive, comprises: Calculating the matching confidence of each package and conversion rule in the metadata and the dynamic rule base; when the highest matching confidence is in a preset threshold range, judging a fuzzy matching state; under the fuzzy matching state, an alternative rule queue is generated according to descending order of confidence, and rules in the queue are executed in sequence until one of the rules is successfully packaged and converted; The execution process of the encapsulation and conversion rule is encapsulated as an atomic operation, and the encapsulation and conversion rule automatically rolls back when any rule fails to execute.
  6. 6. The AI-and-dynamic-rule-based archive informatization management method of claim 1, wherein the multi-modal AI model comprises at least: The document image analysis model is established based on a convolutional neural network and a transducer model and is used for carrying out layout analysis and OCR (optical character recognition) on the new input file in a scanning piece or picture format and converting the extracted data into a structured text; The natural language processing model is used for carrying out entity identification and keyword extraction on the new input file or the structured text in a text format; the text classification model is used for analyzing the density and the up-down Wen Yuyi of the sensitive entities in the archive and automatically outputting the archive sensitivity level.
  7. 7. The AI-and dynamic rule-based archive informatization management method of any one of claims 1-6, wherein when the rights verification of the utilization request is not passed, the method further comprises: And responding to the authority request of the user to the target file, triggering the authority approval workflow, routing the authority request to an approver according to the routing rules in the dynamic rule base, and dynamically updating the authorization item of the corresponding user according to the approval result.
  8. 8. The AI-and dynamic rule-based archive informatization management method of any of claims 1-6, wherein after triggering the utilizing workflow engine, the method further comprises: at least one of a generic watermark, a custom watermark, and a dynamic watermark is obtained, the obtained watermark is added to the archive to be utilized, and/or, And acquiring a custom utilization scene, generating an additional utilization watermark according to the custom utilization scene, and adding the additional utilization watermark into the file to be utilized.
  9. 9. The AI-and dynamic rule-based archive informatization management method of any one of claims 1-6, wherein when the newly entered archive is a paper archive, the method further comprises: And generating an archive chapter according to the metadata of the paper archive, and adding the archive chapter to the archive information package.
  10. 10. An electronic device, the electronic device comprising: at least one processor, and A memory communicatively coupled to the at least one processor, wherein, The memory stores a computer program executable by the at least one processor to enable the at least one processor to perform the AI-and dynamic rule-based archive informatization management method of any one of claims 1-9.

Description

File informatization management method based on AI and dynamic rules and electronic equipment Technical Field The invention relates to the technical field of informatization management, in particular to an archive informatization management method based on AI and dynamic rules and electronic equipment. Background With the continuous perfection of enterprise informatization, solutions for archive informatization management have been gradually applied to various fields. The file informatization management technology converts file information of the traditional medium into digital information through comprehensive application of information technology, digital technology and Internet of things technology, and provides management functions of file input, retrieval and utilization, borrowing management, investigation and statistics and the like. In the prior art, the file informatization management technology is generally based on static rules, and performs file entry, file management and file utilization based on a preset fixed business logic and operation flow. In the file input link, only information of fixed fields can be extracted based on static rules, when a new file format (such as a video or picture format) appears, metadata still needs to be manually registered and supplemented, so that the file input efficiency is low, in the file management link, static authority rules (such as grading authorization or authorization according to user roles) are adopted to classify and graded manage file information, when user attributes change (such as employee adjustment and organization adjustment), authority management relations need to be manually updated by a file manager, authority lag or authority errors easily occur, the file management method is not suitable for scenes with undefined authority levels and more temporary access requirements, in the data utilization and sharing link, the static rules are difficult to realize comprehensive sensitive information desensitization processing, leakage-proof traceability is lacked, the risk of file information leakage and tampering is high, and the security is poor. Disclosure of Invention The invention provides an AI and dynamic rule-based archive informatization management method, which aims to solve the problems of low input efficiency, poor universality of a permission management mechanism, high information leakage risk and the like of the conventional archive informatization management technology. According to one aspect of the invention, an archive informatization management method based on AI and dynamic rules is provided, which comprises the steps of responding to configuration operation of an archive administrator, configuring user groups containing different authorization items, configuring information encapsulation workflow, authority approval workflow and archive utilization workflow based on a predefined dynamic rule base, extracting metadata of a newly entered archive by adopting a pre-trained multi-mode AI model, generating a submitting information packet, driving the information encapsulation workflow according to the metadata, enabling the information encapsulation workflow to search the dynamic rule base, matching and executing corresponding encapsulation and conversion rules based on the metadata, converting the submitting information packet into an archive information packet, storing the archive information packet into an archive base, responding to a utilization request of a user on a target archive, triggering the archive utilization workflow, executing authority verification according to the user groups and the authorization items of the user initiating the request, and executing retrieval and calling on the archive base according to an authority verification result. Optionally, the response to the user request for the target archive triggers the archive utilization workflow, the archive utilization workflow executes authority verification according to a user group to which the user initiating the request belongs and an authorization item thereof, and executes retrieval and calling on the archive according to an authority verification result, wherein the dynamic context portrait at least comprises a semantic intention of the current utilization request, a historical utilization behavior portrait of the user and a risk level of the user group to which the user belongs is built based on the user group to which the user belongs and the authorization item thereof, a dynamic desensitization engine is triggered based on the authority verification result and the dynamic context portrait, and archive calling is executed after the self-adaptive desensitization processing is performed on the retrieved target archive content, wherein the desensitization level of the self-adaptive desensitization processing is inversely related to the risk level of the dynamic context portrait. Optionally, after the archives are searched by utilizing a workflow, the method