Archive warehouse-in and warehouse-out management method and system based on RFID (Radio Frequency Identification)

Abstract

The invention discloses an archive warehouse-in and warehouse-out management method and system based on RFID, and relates to the technical field of warehouse-in and warehouse-out management. The invention discloses an archive warehouse-in and warehouse-out management system based on RFID (Radio Frequency Identification). The archive warehouse-in and warehouse-out management system comprises an archive label distribution module, an archive warehouse-out management module and an archive warehouse-in management module, according to the invention, each archive is provided with a primary archive label and a secondary archive label, and the RFID radio frequency technology is combined, so that the non-contact automatic identification of the identity, state and security level of the archive is realized; an archive confidentiality level classification mechanism is introduced, an image, text and handwriting multi-mode Hash fingerprint technology is combined, accurate content collection and structured abstract storage are carried out on archives which cannot be tampered and are slightly tampered, a reliable basis is provided for subsequent consistency detection, and the archive content tampering detection capability is remarkably enhanced.

Description

Technical Field

[0001] The present invention relates to the technical field of storage-in and storage-out management, and in particular to a file storage-in and storage-out management method and system based on RFID radio frequency identification. Background Art

[0002] Existing methods for managing the entry and exit of archives often rely on manual registration or barcode scanning, making it difficult to achieve refined oversight of the entire archive access process and prevent tampering risks. This is especially true for highly confidential, non-tamperable paper archives, which lack effective automatic identification and consistency detection mechanisms. This can lead to identity confusion, irregular exit approval procedures, and undetectable tampering of returned archive contents. Furthermore, traditional entry and exit management cannot conduct targeted analysis of unusual entry and exit behaviors or repeated access to the same archive, leading to security risks such as archive leaks and unclear responsibilities.

[0003] Therefore, it is necessary to design an intelligent archive in-and-out management method that integrates RFID radio frequency identification with multimodal feature comparison, access behavior modeling, and approval chain anomaly detection to improve the security and traceability of the entire archive process. Summary of the Invention

[0004] The present invention aims to provide a file entry and exit management method and system based on RFID radio frequency identification, so as to improve the security and traceability of the entire file process.

[0005] A file entry and exit management method based on RFID radio frequency identification includes the following steps: Each archived file is assigned a primary file label and a secondary file label. The primary file label includes the confidentiality level, file identification number, and file status. Archived files are divided into tamper-proof files, slightly tamper-proof files, and non-tamper-proof files based on the confidentiality level. Before an archived file is stored, authorized personnel collect data for it and construct the primary file label and secondary file label. When an in-stock file is released from the warehouse, the review and approval responsibility chain and historical review records of the current in-stock file are extracted. Based on the review and approval responsibility chain and historical review records, it is determined whether there is any abnormal review of the in-stock file. If there is no abnormal review, the in-stock file is approved for release and the file status of the corresponding master file tag is changed to release. Otherwise, the in-stock file release approval operation is carried out again. When an in-stock file with the file status of "Out of Stock" is put back into storage, a consistency check is performed on the in-stock file based on the primary file tag and the secondary file tag to obtain the file consistency check result. If the file consistency check result is passed, the file status is changed to "In Stock" and the file is put back into storage. Otherwise, the abnormal file storage operation is started.

[0006] As a preferred technical solution of the present invention, the specific steps of data collection for the archives in the archive and construction of secondary archive labels by authorized warehousing personnel include: The secondary file tag includes the file page number and the corresponding file image hash fingerprint, file text hash fingerprint and file handwriting hash fingerprint; Based on the digital collection of archives in the library page by page, multiple archive images are obtained; hash fingerprints are extracted from the archive images to obtain multiple archive image hash fingerprints; all text information contained in the archive images is identified to obtain multiple archive text hash fingerprints; at the same time, all handwriting features in the archive images are recorded to obtain multiple archive handwriting hash fingerprints.

[0007] As a preferred technical solution of the present invention, the specific steps of determining whether there is abnormal access to archived files based on the access approval responsibility chain and historical access records include: Extract the nth responsible person R based on the review and approval responsibility chain n Historical review and approval records L n , n=1, 2, ..., N; N is the total number of intermediate persons responsible for file access; based on the historical access approval record L n Establish behavioral profile feature vector V n , V n ={f i |i=1, 2, …, I}, where I is the total number of behavioral profile features contained in the behavioral profile feature vector; Set the historical time window of the historical retrieval record. The historical retrieval record contains K retrieval log records D within the historical time window. k , k=1,2,…,K; Based on the behavioral profile feature vector V n Conduct abnormal analysis of approval behavior and obtain the first abnormal review score; Based on the log record D k Conduct abnormal reading behavior analysis to obtain a second abnormal reading score; A comprehensive evaluation is performed based on the first abnormal review score and the second abnormal review score to obtain an abnormal review evaluation result; if the abnormal review evaluation result meets the set abnormal review conditions, an abnormal review phenomenon exists; otherwise, no abnormal review phenomenon exists.

[0008] As a preferred technical solution of the present invention, the specific steps of analyzing abnormal approval behavior and abnormal review behavior include: Based on the behavioral profile feature vector V n The outlier factor of historical behavior is obtained by performing outlier detection on the main file label; the baseline detection of the review and approval responsibility chain is performed based on the pre-trained baseline behavior model to obtain the approval behavior baseline factor; based on the behavior portrait feature vector Vn Establish several historical approval responsibility chains; abstract the several historical approval responsibility chains into a historical approval responsibility chain diagram; determine the similarity between the review approval responsibility chain and the historical approval responsibility chain diagram to obtain the responsibility chain behavior similarity factor; calculate the first abnormal review score based on the historical behavior outlier factor, the approval behavior baseline factor, and the responsibility chain behavior similarity factor; Based on the log record D k Extracting a first access index C1, a second access index C2, and a third access index C3; establishing an access frequency risk scoring function based on the first access index C1, the second access index C2, and the third access index C3 to obtain a second abnormal access score; Among them, the first review index C1 represents the total number of historical reviews, the second review index C2 represents the number of reviews by the same responsible person, and the third review index C3 represents the overlap ratio of the historical approval responsibility chain between the responsible person sets.

[0009] As a preferred technical solution of the present invention, the specific steps of performing consistency detection on the archives in the library based on the primary archive tag and the secondary archive tag include: When an in-stock file that is in the out-of-stock status is re-entered, the file identification number and confidentiality level in the master file tag corresponding to the in-stock file are read; historical consistency detection records are obtained based on the confidentiality level and file identification number; a dynamic detection tolerance model is constructed based on the historical consistency detection records; and image difference thresholds, text change thresholds, and handwriting change thresholds are output based on the dynamic detection tolerance model; Recollect the entry summary data of the archived archive when it is re-entered into the archive; perform multimodal fingerprint comparison between the entry summary data and the secondary archive tag to obtain the archive modification rate set; perform comparison based on the image difference threshold, text modification threshold, handwriting modification threshold and the archive modification rate set to obtain the archive modification consistency result; Extract the signature path of user operation behavior during the outbound process; based on the signature path of user operation behavior and the review and approval responsibility chain file review consistency results; Output the file consistency detection results based on the file modification consistency results and the file retrieval consistency results.

[0010] As a preferred technical solution of the present invention, the primary file tag and the secondary file tag are constructed based on RFID radio frequency technology.

[0011] An RFID-based file entry and exit management system includes: The archive label assignment module includes a label assignment unit, which is used to assign a primary archive label and a secondary archive label to each archive in the library. The primary archive label includes the confidentiality level, archive identity number, and archive status. According to the confidentiality level, the archives in the library are divided into tamper-proof archives, slightly tamper-proof archives, and non-tamper-proof archives. Before the archives are stored in the library, authorized storage personnel collect data for the archives and construct primary archive labels and secondary archive labels. The archive outbound management module includes an outbound verification unit. This unit is used to extract the access approval responsibility chain and historical access records of the current archives when the archives are outbound. Based on the access approval responsibility chain and historical access records, it is determined whether there are any abnormal access phenomena in the archives. If there are no abnormal access phenomena, the archives are approved for outbound access and the archive status of the corresponding master archive tag is changed to outbound access. Otherwise, the outbound access approval operation of the archives is repeated. The archive entry management module includes an entry verification unit; when an archive in the warehouse with an archive status of out of the warehouse is re-entered, it is used to perform a consistency test on the archive in the warehouse based on the primary archive tag and the secondary archive tag to obtain the archive consistency test result; if the archive consistency test result is passed, the archive status is changed to in-warehouse and re-entered; otherwise, the abnormal archive entry operation is started.

[0012] The present invention has the following advantages: 1. The present invention equips each file with a primary file tag and a secondary file tag, and combines RFID radio frequency technology to achieve non-contact automatic identification of the file's identity, status, and security level. It introduces a file confidentiality level classification mechanism and combines image, text, and handwriting multimodal hash fingerprint technology to accurately collect the content and seal the structured summary of non-tamperable and slightly tamperable files, providing a reliable basis for subsequent consistency testing and significantly enhancing the ability to detect file content tampering.

[0013] 2. The present invention combines the review and approval responsibility chain with historical behavior portraits to construct a multi-dimensional abnormal review behavior analysis model, comprehensively evaluates the rationality and frequency risk of the approval path, effectively identifies potential unauthorized access, frequent calls and abnormal collaborative review behaviors, and strengthens the intelligent review capability of outbound approval; in the archive return and warehousing link, through the integration of dynamic detection tolerance model and behavior signature path, a dual comparison of archive change consistency and review rationality is achieved, thereby forming a closed-loop and reliable archive consistency detection mechanism, effectively preventing risks such as archive replacement, concealment and tampering, and overall improving the security, controllability and traceability of high-sensitivity archives in the entire process of entry and exit. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 This is a structural diagram of an RFID-based archive entry and exit management system used in an embodiment of the present invention. DETAILED DESCRIPTION

[0015] In order to enable those skilled in the art to better understand the technical solutions in the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0016] Example 1, a file entry and exit management method based on RFID radio frequency identification, comprising the following steps: Each archived file is assigned a primary file label and a secondary file label. The primary file label includes the confidentiality level, file identification number, and file status. Archived files are divided into tamper-proof files, slightly tamper-proof files, and non-tamper-proof files based on the confidentiality level. Before an archived file is stored, authorized personnel collect data for it and construct the primary file label and secondary file label. The archive identity number is a globally unique identifier assigned to each archive in the library, used to achieve accurate management and traceability of individual archives throughout their life cycle. Its generation is usually completed automatically, using a structured coding method that includes the institution code, year, archive type code, sequence number and check code to ensure the uniqueness, scalability and traceability of the number. It is also bound and stored with all archive operations. The confidentiality level is a key indicator to measure the sensitivity and importance of archive content. It is divided into three levels: tamperable, slightly tamperable and non-tamperable based on the archive source, type field, keyword content and user selection. Among them, non-tamperable archives are subject to the highest level of detection and permission control strategies, such as page-level hash consistency verification and multi-level approval processes. The specific steps for authorized warehousing personnel to collect data for archives in the library and construct sub-archive labels include: The secondary file tag includes the file page number and the corresponding file image hash fingerprint, file text hash fingerprint and file handwriting hash fingerprint; Based on the digital collection of archives in the library page by page, multiple archive images are obtained; hash fingerprints are extracted from the archive images to obtain multiple archive image hash fingerprints; all text information contained in the archive images is identified to obtain multiple archive text hash fingerprints; at the same time, all handwriting features in the archive images are recorded to obtain multiple archive handwriting hash fingerprints; The authorized warehousing personnel first unfold the archives to be stored page by page and place them in a high-resolution image acquisition device, scan each page of content page by page, and generate corresponding digital image files; then, use the image processing module to perform hash fingerprint extraction processing on the collected images, including image preprocessing (such as grayscale, denoising, edge enhancement) and perceptual hashing operations, to generate image hash fingerprints corresponding to each page of the image, which are used for visual similarity comparison in subsequent tampering detection; on this basis, the text area in the image is optically recognized through the integrated OCR engine, the recognition results are structured and stored, and the recognized text content is hashed to form a hash fingerprint of each page of text content, so as to achieve consistency verification of the integrity of the archive text content; at the same time, if there is handwritten information such as signatures and annotations in the image, the handwriting feature extraction module is called to perform stroke segmentation, trajectory extraction, pen pressure modeling and glyph feature extraction on the handwriting area, and further generate a structured description and compressed hash fingerprint of the handwriting.

[0017] When an in-stock file is released from the warehouse, the review and approval responsibility chain and historical review records of the current in-stock file are extracted. Based on the review and approval responsibility chain and historical review records, it is determined whether there is any abnormal review of the in-stock file. If there is no abnormal review, the in-stock file is approved for release and the file status of the corresponding master file tag is changed to release. Otherwise, the in-stock file release approval operation is carried out again. The specific steps for determining whether there are abnormal access phenomena in the archives based on the access approval responsibility chain and historical access records include: Extract the nth responsible person R based on the review and approval responsibility chain n Historical review and approval records L n , n=1, 2, ..., N; N is the total number of intermediate persons responsible for file access; based on the historical access approval record L n Establish behavioral profile feature vector V n , V n ={f i |i=1, 2, …, I}, where I is the total number of behavioral profile features contained in the behavioral profile feature vector; Historical access approval records refer to the system record of all file access activities that a person responsible for access approval has participated in over a certain period of time (such as three months or six months). These records include not only whether the access application was approved, but also information such as the approval time, file ID, applicant information, approval result, and the approval process node. They are used to construct a behavioral profile of the person responsible and determine whether they have abnormal behavior patterns such as a preference for certain types of files or abnormally high approval rates. The review and approval responsibility chain refers to the entire approval process and sequence structure that a certain in-repository file must go through during the out-repository (review) process, i.e., the approval process path defined by the system. Each person in the review and approval responsibility chain has a certain degree of decision-making authority on whether to allow the file to be released from the repository. Historical access records refer to all access records for a currently in-repository file over the past period of time, regardless of whether they are directly related to the current request for release. These records include information such as the access applicant, application time, approval, actual release time, and return time. They are used to determine whether the file has any group behavior anomalies, such as frequent borrowing, group borrowing, or overlapping borrowing paths. When combined with the access approval responsibility chain, potential risks such as overlapping access paths and coordinated borrower behavior can be identified. Set the historical time window of the historical retrieval record. The historical retrieval record contains K retrieval log records D within the historical time window. k , k=1,2,…,K; Based on the behavioral profile feature vector V n Conduct abnormal analysis of approval behavior and obtain the first abnormal review score; Based on the log record D k Conduct abnormal reading behavior analysis to obtain a second abnormal reading score; A comprehensive assessment is performed based on the first and second abnormal review scores to obtain an abnormal review assessment result. If the abnormal review assessment result meets the set abnormal review conditions, an abnormal review phenomenon exists; otherwise, no abnormal review phenomenon exists. The abnormal review conditions are set manually. The first and second abnormal review scores are normalized, and a weighting strategy is set, with weight coefficients α and β, such that α + β = 1. If the weighted result meets the set abnormal review conditions, an abnormal review phenomenon exists.

[0018] The specific steps for analyzing abnormal approval and review behaviors include: Based on the behavioral profile feature vector V n The outlier factor of historical behavior is obtained by performing outlier detection on the main file label; the baseline detection of the review and approval responsibility chain is performed based on the pre-trained baseline behavior model to obtain the approval behavior baseline factor; based on the behavior portrait feature vector V n Establish several historical approval responsibility chains; abstract several historical approval responsibility chains into a historical approval responsibility chain diagram; judge the similarity between the reviewed approval responsibility chain and the historical approval responsibility chain diagram to obtain the responsibility chain behavior similarity factor; calculate the first abnormal review score based on the historical behavior outlier factor, the approval behavior baseline factor and the responsibility chain behavior similarity factor; the first abnormal review score can be obtained by using an average weighted calculation method.

[0019] Based on the historical behavior portrait feature vector of the person responsible for reviewing the file and the main file label of the current file, an anomaly detection algorithm is used to evaluate whether the person's behavior in the current situation significantly deviates from his or her historical behavior pattern, thereby generating a historical behavior outlier factor. The core purpose of performing anomaly detection based on the historical behavior portrait feature vector of the person responsible for reviewing the file and the main file label of the current file is to evaluate whether the person responsible has behavioral anomalies or the risk of role violation in the current approval behavior. The evaluation is performed using an unsupervised anomaly detection algorithm. The higher the value of the historical behavior outlier factor, the more the behavior deviates from the historical pattern and the more suspicious it is. As the first-level risk identification mechanism, the historical behavior outlier factor can provide timely warnings when the approval behavior of the person responsible first shows a pattern outside of habit. The historical behavior outlier factor complements individual behavior deviations that are difficult to detect due to review frequency and path structure, such as someone suddenly approving a certain type of highly confidential file in a concentrated manner.

[0020] Using a pre-trained benchmark behavior model, the current access and approval responsibility chain is structurally matched with common responsibility chain paths in normal archive release processes to determine whether the path is a high-frequency, regular path or a rare path. Based on this, a benchmark factor for approval behavior is generated. The pre-trained benchmark behavior model collects a large number of real access and approval responsibility chain paths from historical archive release processes as training samples. Each responsibility chain is represented as a directed path or graph structure, and features such as node sequence, role hierarchy, and frequency of responsible individuals are extracted. The responsibility chain is converted into a learnable vector representation using graph structure encoding methods. The model is trained using unsupervised learning methods such as cluster analysis or autoencoders, thereby establishing a benchmark space for high-frequency, normal responsibility chain patterns. In actual operation, the responsibility chain structure corresponding to the current release application is input into the model, and a similarity match is performed with the benchmark space. A benchmark match score is output as the approval behavior benchmark factor. The actual function of this factor is to measure whether the current responsibility chain path belongs to the common and reasonable approval process known within the system, thereby identifying the risks of unauthorized overreach, process circumvention, or potential illegal linkages caused by rare or atypical paths. This helps to identify structural release anomalies in advance and trigger audit mechanisms.

[0021] Retrieve multiple responsibility chain paths that the responsible person participated in in the past approval behavior, construct them into a historical responsibility chain set, and abstract them into a directed responsibility chain graph. Calculate the structural similarity between the current approval chain and the historical graph through graph embedding or graph edit distance algorithm to obtain the responsibility chain behavior similarity factor. The actual role of the responsibility chain behavior similarity factor is to measure whether the current review of the approval responsibility chain conforms to the responsibility chain structure characteristics that the responsible person has participated in the past. It is mainly used to identify whether there are suspicious behaviors such as abnormal insertions or role changes in the approval path of the responsible person. By constructing multiple responsibility chain paths that the responsible person has participated in the past into a historical responsibility chain set and abstracting them into a graph, After the directed graph is constructed, the current approval chain can be structurally compared with the historical responsibility chain graph based on algorithms such as graph embedding or graph edit distance to determine their structural similarity. If the similarity is low, it means that the current approval path is significantly different from the previous behavior of the responsible person, which may imply that his current behavior deviates from the conventional role positioning, he is temporarily added to an uncommon process, or he is in an unusual path across departments. Therefore, the responsibility chain behavior similarity factor can assist in identifying non-explicit risk behaviors such as process crossing, approval chain tampering, and regulatory evasion from a structural dimension. It is an important supplement to the individual behavior portrait and path benchmark model, and helps to improve the depth and reliability of approval anomaly detection.

[0022] Based on the log record D k Extracting a first access index C1, a second access index C2, and a third access index C3; establishing an access frequency risk scoring function based on the first access index C1, the second access index C2, and the third access index C3 to obtain a second abnormal access score; The construction process of the access frequency risk scoring function aims to quantify multiple frequency characteristics reflecting archive access behavior into a comprehensive score, which is used to measure whether a certain archive has abnormally frequent or abnormally related usage behavior within a specific time window. The access frequency risk scoring function can be established using a weighted calculation method. The weights can be set through supervised learning training of historical access anomaly samples or expert experience to reflect the relative importance of each indicator in risk assessment. Among them, the first review index C1 represents the total number of historical reviews, the second review index C2 represents the number of reviews by the same responsible person, and the third review index C3 represents the overlap ratio of the historical approval responsibility chain between the responsible person groups; The total number of accesses is used to assess whether the archived files are accessed at an abnormal frequency; the number of accesses by the same responsible person is used to determine whether there is suspicious behavior of a specific responsible person frequently accessing specific files; the overlap ratio of the historical approval responsibility chain is used to identify whether there is clustered collaborative behavior among approvers or potential interest relationships.

[0023] When an in-stock file with the status of "Out of Stock" is re-stocked, a consistency check is performed on the in-stock file based on the primary and secondary file tags to obtain the file consistency check result. If the file consistency check result is passed, the file status is changed to "In Stock" and the file is re-stocked. Otherwise, the abnormal file re-stocking operation is initiated. The specific steps for consistency checking of archives in the library based on primary archive tags and secondary archive tags include: When an in-stock file that is in the out-of-stock status is re-entered, the file identification number and confidentiality level in the master file tag corresponding to the in-stock file are read; historical consistency detection records are obtained based on the confidentiality level and file identification number; a dynamic detection tolerance model is constructed based on the historical consistency detection records; and image difference thresholds, text change thresholds, and handwriting change thresholds are output based on the dynamic detection tolerance model; Recollect the entry summary data of the archived archive when it is re-entered into the archive; perform multimodal fingerprint comparison between the entry summary data and the secondary archive tag to obtain the archive modification rate set; perform comparison based on the image difference threshold, text modification threshold, handwriting modification threshold and the archive modification rate set to obtain the archive modification consistency result; Extract the signature path of user operation behavior during the outbound process; based on the signature path of user operation behavior and the review and approval responsibility chain file review consistency results; Output the file consistency detection results based on the file modification consistency results and the file retrieval consistency results.

[0024] In the process of re-entering the archive after it has been released from the warehouse, consistency testing is a key step to ensure that the original appearance of the archive has not been tampered with. First, the archive identity number and confidentiality level in the master file tag corresponding to the current archive are read. The archive identity number is used to uniquely locate the historical detection behavior record of the archive, and the confidentiality level determines the strictness of the comparison tolerance strategy. Based on these two pieces of information, its historical consistency detection records (including summary features of each entry, past comparison differences, etc.) are retrieved, and based on the statistical characteristics of historical detection deviations, a dynamic detection tolerance model corresponding to the individual archive is constructed. This model can analyze historical difference data through Bayesian estimation or kernel density estimation to output The reasonable modification tolerance threshold of the archive in different modalities; after obtaining the image difference threshold, text modification threshold and handwriting modification threshold, the storage summary data collection operation of the current archive is performed, that is, the archive is re-scanned or its digitized image is read page by page, and its text content is extracted using OCR technology (optical character recognition), and the handwriting features are extracted using the handwriting recognition model; based on the image hash algorithm, text hash and handwriting feature encoding, a multimodal summary fingerprint is compared with the original features in the secondary archive label, and the difference rate of each page in the three modalities is calculated. The final output is a set of archive modification rates, recording the image modification rate, text modification rate and handwriting modification rate; Based on the threshold value corresponding to the file and the set of change rates, each item is compared to determine whether the difference is within the acceptable range. Combined with the results, it is determined whether there are any abnormal changes in graphics, text, or handwriting. The consistency results of the file changes for the batch of files are obtained and the output data is in the form of scores. The signature path of the user's operation behavior recorded during the release of the file is called up, that is, the behavior chain generated by the user when accessing the file, such as the time the file was accessed, whether the user immediately applied to access other related files after accessing the file, etc., and analyzed through behavior trajectory compression and path modeling to determine whether it is consistent with the normal access pattern. Combined with the responsibility chain of the release approval, it is evaluated whether there is any unreasonable behavior path in the file access behavior, unauthorized access, or inconsistency between the access purpose and the usage behavior, and the file access consistency result is comprehensively output; The consistency results of archive changes and the consistency results of archive retrieval are weighted and integrated, and the final archive consistency test results of the archive are generated based on preset rules or machine learning models (such as logistic regression or decision trees). If the result is passed, the archive status is updated from "out of storage" to "in storage", and the legal storage operation is completed; if the result is failed, it will automatically enter the review process, and manual judgment will be made whether there is tampering, replacement, substitution or other violations, thereby effectively realizing security verification and risk identification when the out-of-stock archives are returned.

[0025] The main file tag and secondary file tag are constructed based on RFID radio frequency technology. In the file entry and exit management system based on RFID radio frequency technology, both the main file tag and the secondary file tag can realize digital identification and automatic sensing recognition through RFID chips, which runs through the entire process of file identity calibration, entry and exit tracking, security verification and behavior monitoring; during the file entry stage, the main file tag is embedded with an RFID chip to record core identity information such as file identity number, confidentiality level, file status, etc. The RFID tag is fixed to the front or back cover of the file by attaching or embedding, and has the characteristics of not being easily removed or replaced, ensuring the long-term uniqueness and traceability of the file's identity; in the process of file entry and exit management, RFID realizes contactless batch identification and precise positioning; when the file is retrieved or returned, the file's identity number and status are identified in real time through RFID reading equipment (such as bookshelf antennas, channel doors, desktop inventory meters, etc.), and its entry and exit records are automatically updated to avoid manual scanning errors; when the file enters the reading area, its presence and stay in the reading area can be sensed by the regional RFID base station, realizing the node record of the reading behavior; if combined with the positioning module, it can even record the file's reading path in a certain row of bookshelves or a certain workstation, which is used to assist in restoring the user's reading behavior chain.

[0026] Example 2, a file entry and exit management system based on RFID radio frequency identification, see Figure 1 Shown, including: The archive label assignment module includes a label assignment unit, which is used to assign a primary archive label and a secondary archive label to each archive in the library. The primary archive label includes the confidentiality level, archive identity number, and archive status. According to the confidentiality level, the archives in the library are divided into tamper-proof archives, slightly tamper-proof archives, and non-tamper-proof archives. Before the archives are stored in the library, authorized storage personnel collect data for the archives and construct primary archive labels and secondary archive labels. The archive outbound management module includes an outbound verification unit. This unit is used to extract the access approval responsibility chain and historical access records of the current archives when the archives are outbound. Based on the access approval responsibility chain and historical access records, it is determined whether there are any abnormal access phenomena in the archives. If there are no abnormal access phenomena, the archives are approved for outbound access and the archive status of the corresponding master archive tag is changed to outbound access. Otherwise, the outbound access approval operation of the archives is repeated. The archive entry management module includes an entry verification unit; when an archive in the warehouse with an archive status of out of the warehouse is re-entered, it is used to perform a consistency test on the archive in the warehouse based on the primary archive tag and the secondary archive tag to obtain the archive consistency test result; if the archive consistency test result is passed, the archive status is changed to in-warehouse and re-entered; otherwise, the abnormal archive entry operation is started.

[0027] It should be understood that those skilled in the art may make improvements or modifications based on the above description, and all such improvements and modifications shall fall within the scope of protection of the appended claims. Any portion of this specification not described in detail is prior art known to those skilled in the art.

Claims

1. A file entry and exit management method based on RFID radio frequency identification, characterized in that: The following steps are involved: Each archived file is assigned a primary file label and a secondary file label. The primary file label includes the confidentiality level, file identification number, and file status. Archived files are divided into tamper-proof files, slightly tamper-proof files, and non-tamper-proof files based on the confidentiality level. Before an archived file is stored, authorized personnel collect data for it and construct the primary file label and secondary file label. When an in-stock file is released from the warehouse, the review and approval responsibility chain and historical review records of the current in-stock file are extracted. Based on the review and approval responsibility chain and historical review records, it is determined whether there is any abnormal review of the in-stock file. If there is no abnormal review, the in-stock file is approved for release and the file status of the corresponding master file tag is changed to release. Otherwise, the in-stock file release approval operation is carried out again. When an in-stock file with the file status of "Out of Stock" is put back into storage, a consistency check is performed on the in-stock file based on the primary file tag and the secondary file tag to obtain the file consistency check result. If the file consistency check result is passed, the file status is changed to "In Stock" and the file is put back into storage. Otherwise, the abnormal file storage operation is started.

2. The RFID-based file entry and exit management method according to claim 1, characterized in that: The specific steps for authorized warehousing personnel to collect data for archives in the library and construct sub-archive labels include: The secondary file tag includes the file page number and the corresponding file image hash fingerprint, file text hash fingerprint and file handwriting hash fingerprint; Based on the digital collection of archives in the library page by page, multiple archive images are obtained; hash fingerprints are extracted from the archive images to obtain multiple archive image hash fingerprints; all text information contained in the archive images is identified to obtain multiple archive text hash fingerprints; at the same time, all handwriting features in the archive images are recorded to obtain multiple archive handwriting hash fingerprints.

3. The RFID-based file entry and exit management method according to claim 2, characterized in that: The specific steps for determining whether there are abnormal access phenomena in the archives based on the access approval responsibility chain and historical access records include: Extract the nth responsible person R based on the review and approval responsibility chain n Historical review and approval records of L n , n=1, 2, ..., N; N is the total number of intermediate persons responsible for file access; based on the historical access approval record L n Establish behavioral profile feature vector V n , V n ={f i |i=1, 2, …, I}, where I is the total number of behavioral profile features contained in the behavioral profile feature vector; Set the historical time window of the historical retrieval record. The historical retrieval record contains K retrieval log records D within the historical time window. k , k=1,2,…,K; Based on the behavioral profile feature vector V n Conduct abnormal analysis of approval behavior and obtain the first abnormal review score; Based on the log record D k Conduct abnormal reading behavior analysis to obtain a second abnormal reading score; A comprehensive evaluation is performed based on the first abnormal review score and the second abnormal review score to obtain an abnormal review evaluation result; if the abnormal review evaluation result meets the set abnormal review conditions, an abnormal review phenomenon exists; otherwise, no abnormal review phenomenon exists.

4. The RFID-based file entry and exit management method according to claim 3 is characterized in that: The specific steps for analyzing abnormal approval and review behaviors include: Based on the behavioral profile feature vector V n The outlier factor of historical behavior is obtained by performing outlier detection on the main file label; the baseline detection of the review and approval responsibility chain is performed based on the pre-trained baseline behavior model to obtain the approval behavior baseline factor; based on the behavior portrait feature vector V n Establish several historical approval responsibility chains; abstract the several historical approval responsibility chains into a historical approval responsibility chain diagram; determine the similarity between the review approval responsibility chain and the historical approval responsibility chain diagram to obtain the responsibility chain behavior similarity factor; calculate the first abnormal review score based on the historical behavior outlier factor, the approval behavior baseline factor, and the responsibility chain behavior similarity factor; Based on the log record D k Extracting a first access index C1, a second access index C2, and a third access index C3; establishing an access frequency risk scoring function based on the first access index C1, the second access index C2, and the third access index C3 to obtain a second abnormal access score; Among them, the first review index C1 represents the total number of historical reviews, the second review index C2 represents the number of reviews by the same responsible person, and the third review index C3 represents the overlap ratio of the historical approval responsibility chain between the responsible person sets.

5. The RFID-based file entry and exit management method according to claim 4 is characterized in that: The specific steps for consistency checking of archives in the library based on primary archive tags and secondary archive tags include: When an in-stock file that is in the out-of-stock status is re-entered, the file identification number and confidentiality level in the master file tag corresponding to the in-stock file are read; historical consistency detection records are obtained based on the confidentiality level and file identification number; a dynamic detection tolerance model is constructed based on the historical consistency detection records; and image difference thresholds, text change thresholds, and handwriting change thresholds are output based on the dynamic detection tolerance model; Recollect the entry summary data of the archived archive when it is re-entered into the archive; perform multimodal fingerprint comparison between the entry summary data and the secondary archive tag to obtain the archive modification rate set; perform comparison based on the image difference threshold, text modification threshold, handwriting modification threshold and the archive modification rate set to obtain the archive modification consistency result; Extract the signature path of user operation behavior during the outbound process; based on the signature path of user operation behavior and the consistency result of reviewing the approval responsibility chain archive; Output the file consistency detection results based on the file modification consistency results and the file retrieval consistency results.

6. The RFID-based file entry and exit management method according to claim 5, characterized in that: The primary file tag and the secondary file tag are constructed based on RFID radio frequency technology.

7. An RFID-based file entry and exit management system, characterized in that: The system is a file entry and exit management method based on RFID radio frequency identification as described in any one of claims 1 to 6, comprising: The archive label assignment module includes a label assignment unit, which is used to assign a primary archive label and a secondary archive label to each archive in the library. The primary archive label includes the confidentiality level, archive identity number, and archive status. According to the confidentiality level, the archives in the library are divided into tamper-proof archives, slightly tamper-proof archives, and non-tamper-proof archives. Before the archives are stored in the library, authorized storage personnel collect data for the archives and construct primary archive labels and secondary archive labels. The archive outbound management module includes an outbound verification unit. This unit is used to extract the access approval responsibility chain and historical access records of the current archives when the archives are outbound. Based on the access approval responsibility chain and historical access records, it is determined whether there are any abnormal access phenomena in the archives. If there are no abnormal access phenomena, the archives are approved for outbound access and the archive status of the corresponding master archive tag is changed to outbound access. Otherwise, the outbound access approval operation of the archives is repeated. The archive entry management module includes an entry verification unit; when an archive in the warehouse with an archive status of out of the warehouse is re-entered, it is used to perform a consistency test on the archive in the warehouse based on the primary archive tag and the secondary archive tag to obtain the archive consistency test result; if the archive consistency test result is passed, the archive status is changed to in-warehouse and re-entered; otherwise, the abnormal archive entry operation is started.

Images