A digital flexible system and method for document control

Abstract

The present invention relates to the technical field of document control, and specifically provides a digital flexible system and method for document control. The system includes a semantic construction module, an event-driven module, an access control engine module, and a behavior analysis and response module. In the present invention, by dynamically adjusting the real-time analysis of document update and operation events and their dependencies, the usage status and changes of files can be captured more accurately, greatly improving the flexibility and response speed of document management. By relying on the analysis and pattern recognition of user behavior, the permission adjustment is more precise, effectively enhancing the security and compliance of documents, significantly improving the overall efficiency and reliability of the file management system, being able to better adapt to the rapidly changing business environment, accurately tracking abnormal behaviors and quickly responding, ensuring a high level of security and control, reducing the possibility of security vulnerabilities, and thus avoiding the risks of data leakage and abuse.

Description

Technical Field

[0001] The present invention relates to the technical field of document control, and particularly to a digital flexible system and method for document control. Background Art

[0002] The technical field of document control involves using digital tools and systems to manage and control files and their related access rights, version control, and audit trails. It usually includes an integrated solution for file storage, retrieval, security, and sharing functions, aiming to improve the efficiency and security of information flow in an organization. Through automated processing programs, the document control system can ensure the consistency and accuracy of documents, reduce human errors, enhance data protection, usually support multiple document types, including text files, images, and spreadsheets, and enable seamless document sharing and collaboration between multiple departments and geographical locations.

[0003] Among them, the digital flexible system for document control is a technical solution specifically designed to optimize and simplify the document management process. The main purpose of the system is to provide more flexible document management functions through digital means, such as real-time editing permission adjustment, automatic version update, and advanced access control mechanisms. The aim is to enhance file security, improve the efficiency of document processing, and support the document sharing and collaboration needs in a remote working environment, so as to adapt to the rapidly changing business environment and increasingly complex compliance requirements.

[0004] Existing document control technologies show limitations in dealing with large-scale real-time data and advanced access control. Especially in the file sharing requirements across departments or geographical locations, they are often not agile enough in real-time editing permission adjustment and version control, making it difficult to cope with rapidly changing work needs, and restricting the adaptability of the system when remote working modes become popular. In addition, existing systems are usually not deep enough in user behavior analysis and abnormal behavior identification, affecting their ability to prevent data leakage and abuse, which may lead to processing delays, affect the operational efficiency and information security of the organization, and increase the risk of data leakage. Summary of the Invention

[0005] The purpose of the present invention is to solve the drawbacks existing in the prior art, and to propose a digital flexible system and method for document control.

[0006] To achieve the above purpose, the present invention adopts the following technical solutions: A digital flexible system for document control includes:

[0007] The semantic construction module extracts keywords from the document, identifies key phrases, draws a dependency graph through connections between entities, monitors document updates, dynamically adjusts the dependency relationship according to the updates, reveals the real-time connections between documents, and obtains the document semantic dependency graph;

[0008] The event-driven module captures all operation events of the document based on the document semantic dependency graph, updates the event log using the captured data, responds to the events, records the event processing process, and generates event response records;

[0009] Based on the user data in the event response record, the access control engine module conducts data statistical analysis, determines the user behavior pattern, updates the access permissions in real time according to the determination result, synchronously updates the access log, and obtains the document permission adjustment record;

[0010] The behavior analysis and response module analyzes the document permission adjustment record, extracts the user access pattern, conducts trend analysis of the user behavior data, detects the deviation from the standard pattern according to the analysis result, identifies abnormal behaviors and responds to the abnormal behaviors, formulates and implements security measures, and obtains the document security control record.

[0011] As a further solution of the present invention, the steps for obtaining the document semantic dependency graph are as follows:

[0012] Separate the text content and structure, remove irrelevant characters and stop words, and use the formula

[0013]

[0014] Calculate the importance score TFIDF of word t imp (t, d), generate a set of keywords and key phrases, where ε is a small amount for fine-tuning the term frequency to avoid the zero-frequency problem, δ is an adjustment amount for the total number of documents, d represents the document, N represents the total number of documents, DF(t) represents the number of documents including word t, and TF(t, d) is the term frequency of word t in document d;

[0015] Based on the set of keywords and key phrases, identify the core entities in the document, and use the formula

[0016]

[0017] Calculate the dependency strength R between entities i and j adj (i, j), generate a preliminary document dependency graph, where α and β are regularization parameters used to balance the influence of entities on the result, CO(i, j) is the number of times entities i and j co-occur in the same document, and DN(i, j) is the number of documents including entities i and j;

[0018] Conduct document monitoring, dynamically update and adjust the structure of the preliminary document dependency graph, and use the formula

[0019]

[0020] Calculate the importance score TFIDF of word t after the change dyn (t, d new), generate a dynamically adjusted document dependency graph, where γ and η are adjustment parameters used to adjust the impact of document updates, d new represents the updated document, DC(t) represents the number of document updates including word t, TF(t, d new ) is the updated document d new the term frequency of word t in the document;

[0021] Based on the dynamically adjusted document dependency graph, use the formula,

[0022]

[0023] calculate the centrality Centr fin (v) of the document node v, and output the document semantic dependency graph, where ξ is a balance coefficient used to adjust the out-degree impact, v represents the document node, NB(v) represents other document nodes directly connected to the document node v, and DO(u) represents the number of documents directly connected to document u.

[0024] As a further solution of the present invention, the step of obtaining the event response record is:

[0025] Based on the document semantic dependency graph, capture the operation events of the document, and use the formula,

[0026]

[0027] calculate the event code LogEntry(E i ) of the i-th event, and generate a preliminary event log, where α E , β E , γ E and δ E are adjustment coefficients used to optimize the accuracy and flexibility of the record, E i represents the i-th event, T i represents the event time, U i represents the user ID, A i represents the user operation type number;

[0028] Based on the preliminary event log, add the server timestamp and verify the user identity, and use the formula,

[0029]

[0030] and

[0031]

[0032] calculate the adjusted timestamp Timestamp(E i ) and the verified log entry value Auth(U i) to generate an event log with timestamps and authentication, where Δt represents the deviation between the system time and the event time, ε E , ζ U and η U are adjustment coefficients used to optimize the accuracy of timestamp application and user authentication, and key is the encryption key;

[0033] Encrypt the event log with timestamps and authentication using the formula,

[0034]

[0035] Calculate the encrypted log entry value Encrypt(AU i ), to obtain the encrypted event log, where AU i represents the authenticated log entry, ξ L and λ L are coefficients used to adjust the encryption operation;

[0036] Based on the encrypted event log, upload it to the central database for storage and record, and output the event response record.

[0037] As a further solution of the present invention, the determination steps of the user behavior pattern are:

[0038] Based on the event response record, extract key user behavior data, including access path, click frequency input data, using the formula,

[0039]

[0040] Calculate the weighted value of user i in behavior j to generate weighted user behavior data, where a ik represents the behavior data of user i in the kth activity of behavior j, m represents the total number of activities, and w k represents the weight coefficient of the kth activity;

[0041] Use the weighted user behavior data to normalize the data, using the formula,

[0042]

[0043] Calculate the normalized score Z ij of the user behavior data to generate normalized user behavior data, where λ j is the normalization parameter of behavior j, and μ j and σ j represent the average value and standard deviation of behavior j respectively;

[0044] Analyze the normalized user behavior data, using the formula,

[0045]

[0046] Calculate the behavior pattern group Q assigned to user i, and generate a user behavior pattern determination result, where ρ k is the adjustment coefficient of cluster k, and G k represents the set of users assigned to the k-th group, and Z i represents the normalized behavior data vector of user i, and C k represents the center vector of the k-th group, and ∥Z i -C k ∥ represents the modulus length.

[0047] As a further solution of the present invention, the step of obtaining the document permission adjustment record is:

[0048] According to the user behavior pattern determination result, set the access permission level for each behavior pattern group, and use the formula

[0049]

[0050] Calculate the access permission level P representing the k-th group k , and generate a group access permission configuration, where V k is the behavior activity coefficient of group k, and L k is the behavior center value of group k, reflecting the behavior pattern of the group, and H k is the historical compliance score, and θ k is the permission threshold of group k;

[0051] According to the group access permission configuration, update the document access permission database, and use the formula

[0052]

[0053] Calculate the permission update status value ZT, and generate a real-time access permission update record, where ε k is the sensitivity parameter, used to adjust the sensitivity of the update amplitude, and δ k is the update threshold, used to fine-tune the permission change;

[0054] Based on the real-time access permission update record, record the permission adjustment content of each user, and use the formula

[0055]

[0056] Calculate the ratio B of the user's last login time to the current time, and generate a document permission adjustment record, where L u represents the last login time of user u, T is the current time, and τ is the time stamp of the permission update.

[0057] As a further solution of the present invention, the analysis steps of the user behavior data trend are as follows:

[0058] Based on the document permission adjustment record, obtain the user ID, document ID, access time, and permission status, and use the formula

[0059]

[0060] Calculate the total number of accesses D of user uid uid , generate user access count data, where v uid represents the number of accesses of user uid to document d, and n is the number of documents accessed;

[0061] Based on the user access count data, identify the user's access pattern, and use the formula

[0062]

[0063] Calculate the weighted sum M of the number of accesses per time unit of the user uid , determine the user access frequency and preference, and generate the user access pattern analysis result, where D uid (t) is the number of accesses of the user at time t;

[0064] Based on the user access pattern analysis result, analyze the user behavior trend, and use the formula

[0065]

[0066] Calculate the quantitative value T of the change trend of user behavior uid , identify the changes and periodicity of the behavior pattern, and generate the user behavior trend analysis result, where is the average value of M uid .

[0067] As a further solution of the present invention, the acquisition steps of the document security control record are as follows:

[0068] According to the user behavior trend analysis result, use the formula

[0069] Δ uid =|T uid -S|

[0070] Calculate the deviation value Δ from the normal behavior uid , generate the deviation detection result, where S is the preset standard behavior pattern value;

[0071] Evaluate the deviation detection result to determine whether it exceeds the preset threshold Y. If Δ uidIf it is >Y, it is marked as abnormal; otherwise, it is marked as normal, and the abnormal behavior identification result is generated.

[0072] According to the abnormal behavior identification result, safety measures are formulated and implemented, and the document security control record is output.

[0073] A digital flexible method for document control includes the following steps:

[0074] S1: Separate the text content and structure, calculate the word importance score, identify the core entities in the document, conduct document monitoring, dynamically update and adjust the structure of the preliminary document dependency graph, and output the document semantic dependency graph.

[0075] S2: Based on the document semantic dependency graph, capture the operation events of the document, add the server timestamp, verify the user identity, implement encryption processing and upload it to the central database for storage record, and output the event response record.

[0076] S3: Based on the event response record, extract the key user behavior data, calculate the user behavior weighted value, standardize the data, conduct behavior pattern grouping, and generate the user behavior pattern determination result.

[0077] S4: According to the user behavior pattern determination result, set the access permission level for each behavior pattern group, update the document access permission database, record the permission adjustment content of each user, and generate the document permission adjustment record.

[0078] S5: Based on the document permission adjustment record, calculate the total access times of the user, identify the access pattern of the user, determine the user access frequency and preference, analyze the user behavior trend, identify the changes and periodicity of the behavior pattern, and generate the user behavior trend analysis result.

[0079] S6: According to the user behavior trend analysis result, calculate the deviation value from the normal behavior, determine whether it exceeds the preset threshold and mark the user behavior, formulate and implement safety measures, and output the document security control record.

[0080] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0081] In the present invention, through real-time analysis of document update and operation events and their dynamic adjustments of dependencies, the usage status and changes of files can be accurately captured, greatly enhancing the flexibility and response speed of document management. By relying on the analysis of user behavior and pattern recognition, permission adjustments can be made more precise, effectively enhancing the security and compliance of documents, significantly improving the overall efficiency and reliability of the file management system, being able to better adapt to the rapidly changing business environment, accurately tracking abnormal behaviors and quickly responding, ensuring a high level of security and control, reducing the possibility of security vulnerabilities, and avoiding the risks of data leakage and abuse. BRIEF DESCRIPTION OF THE DRAWINGS

[0082] Figure 1 It is a system flow chart of the present invention;

[0083] Figure 2 It is a flow chart for obtaining the semantic dependency graph of the documents of the present invention;

[0084] Figure 3 It is a flow chart for obtaining the event response records of the present invention;

[0085] Figure 4 It is a flow chart for determining the user behavior pattern of the present invention;

[0086] Figure 5 It is a flow chart for obtaining the document permission adjustment records of the present invention;

[0087] Figure 6 It is a flow chart for analyzing the trend of user behavior data of the present invention;

[0088] Figure 7 It is a flow chart for obtaining the document security control records of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0089] In order to make the objectives, technical solutions and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0090] In the description of the present invention, it should be understood that the orientation or positional relationships indicated by the terms "length", "width", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. are based on the orientation or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus should not be construed as limiting the present invention. In addition, in the description of the present invention, "a plurality of" means two or more unless otherwise specifically defined.

[0091] Please refer to Figure 1 , a digital flexible system for document control includes:

[0092] The semantic construction module extracts keywords from the document, identifies key phrases, draws a dependency graph through connections between entities, monitors document updates, dynamically adjusts the dependency relationship according to the updates, reveals the real-time connections between documents, and obtains the document semantic dependency graph;

[0093] The event-driven module, based on the document semantic dependency graph, captures all operation events of the document, updates the event log using the captured data, responds to the events, records the event processing process, and generates an event response record;

[0094] The access control engine module, based on the user data in the event response record, conducts data statistical analysis, determines the user behavior pattern, updates the access permission in real time according to the determination result, synchronously updates the access log, and obtains the document permission adjustment record;

[0095] The behavior analysis and response module analyzes the document permission adjustment record, extracts the user access pattern, conducts a trend analysis of the user behavior data, detects the deviation from the standard pattern according to the analysis result, identifies abnormal behaviors and responds to the abnormal behaviors, formulates and implements security measures, and obtains the document security control record.

[0096] The document semantic dependency graph includes keywords, key phrases, dependency relationship records, and entity association results. The event response record includes operation event capture records, event log update records, and event processing records. The document permission adjustment record includes permission update records, user behavior analysis results, and access log synchronization details. The document security control record includes behavior trend analysis results, abnormal behavior identification results, and security measure details.

[0097] Please refer to Figure 2 , the steps for obtaining the document semantic dependency graph are as follows:

[0098] Separate the text content and structure, remove irrelevant characters and stop words, and use the formula,

[0099]

[0100] Calculate the importance score TFIDF imp (t, d) of word t to generate a set of keywords and key phrases, where ∈ is a small amount for fine-tuning the word frequency to avoid the zero-frequency problem, δ is the adjustment amount of the total number of documents, d represents the document, N represents the total number of documents, DF(t) represents the number of documents including word t, and TF(t, d) is the word frequency of word t in document d;

[0101] Based on the set of keywords and key phrases, identify the core entities in the document, and use the formula,

[0102]

[0103] Calculate the dependency strength R of entities i and j adj (i, j), and generate a preliminary document dependency graph, where α and β are regularization parameters used to balance the influence of entities on the result, CO(i, j) is the number of times entities i and j co-occur in the same document, and DN(i, j) is the number of documents including entities i and j;

[0104] Perform document monitoring and dynamically update and adjust the structure of the preliminary document dependency graph using the formula

[0105]

[0106] Calculate the importance score TFIDF of the changed word t dyn (t, d new ), and generate a dynamically adjusted document dependency graph, where γ and η are adjustment parameters used to adjust the influence of document updates, d new represents the updated document, DC(t) represents the number of document updates including word t, TF(t, d new ) is the word frequency of word t in the updated document d new ;

[0107] Based on the dynamically adjusted document dependency graph, use the formula

[0108]

[0109] Calculate the centrality Centr fin (v) of the document node v and output the document semantic dependency graph, where ξ is a balance coefficient used to adjust the influence of out-degree, v represents the document node, NB(v) represents other document nodes directly connected to the document node v, and DO(u) represents the number of documents directly connected to the document u.

[0110] N: The total number of documents, assumed to be 1000.

[0111] DF(t): The number of documents containing word t, assumed to be 50.

[0112] ∈: A small amount for fine-tuning the word frequency, assumed to be 0.01.

[0113] δ: The adjustment amount of the total number of documents, assumed to be 5.

[0114] Assume that in document d, the word frequency TF(t, d) of word t is 10, and calculate according to the formula:

[0115]

[0116] The calculated TFIDF value is 29.8298. The value represents the importance of word t in document d, and a higher value means higher importance.

[0117] CO(i,j): The number of times entities i and j co-occur in the same document, assumed to be 20 times.

[0118] DN(i,j): The number of documents containing entities i and j, assumed to be 10.

[0119] α: Regularization parameter, assumed to be 1.

[0120] β: Regularization parameter, assumed to be 2.

[0121] Substituting into the formula, we get:

[0122]

[0123] The obtained dependence strength is 1.75. The value indicates the degree of association between entities i and j in the document set, and a larger value means a higher degree of association.

[0124] DC(t): The number of document updates involving word t, assumed to be 5.

[0125] γ: Adjustment parameter, assumed to be 0.05.

[0126] η: Adjustment parameter, assumed to be 3.

[0127] Assume TF(t,d new ) is 15, then:

[0128]

[0129] The obtained dynamically adjusted TFIDF value is 42.8925, indicating an increase in the importance of word t in the updated document.

[0130] DO(u): The number of documents directly connected to document u, assumed to be 30.

[0131] ξ: Balance coefficient, assumed to be 0.1.

[0132] Assume R fin (u,v) averages 1.8. Considering three nodes, then:

[0133]

[0134] The obtained centrality analysis value is 0.984, indicating the centrality of document v in the entire document network and reflecting its influence and importance in the document set.

[0135] Please refer to Figure 3 , and the steps for obtaining event response records are:

[0136] Capture the operation events of the document based on the document semantic dependency graph, and use the formula

[0137]

[0138] Calculate the event code LogEntry(E i ) of the i-th event to generate a preliminary event log, where α E , β E , γ E and δ E are adjustment coefficients used to optimize the accuracy and flexibility of the record, E i represents the i-th event, T i represents the event time, U i represents the user ID, and A i represents the user operation type number;

[0139] Based on the preliminary event log, add the server timestamp and verify the user identity, using the formula

[0140]

[0141] and

[0142]

[0143] Calculate the adjusted timestamp Timestamp(E i ) and the verified log entry value Auth(U i ) to generate an event log with timestamp and authentication, where Δt represents the deviation between the system time and the event time, and ε E , ζ U and η U are adjustment coefficients used to optimize the timestamp application and user verification accuracy, and key is the encryption key;

[0144] Encrypt the event log with timestamp and authentication, using the formula

[0145]

[0146] Calculate the encrypted log entry value Encrypt(AU i ) to obtain the encrypted event log, where AU i represents the verified log entry, and ξ L and λ L are coefficients used to adjust the encryption operation;

[0147] Based on the encrypted event log, upload it to the central database for storage and recording, and output the event response record.

[0148] Hypothesis:

[0149] T i = 1609459200 (Unix timestamp, 00:00 on January 1, 2021)

[0150] U i = 1001

[0151] A i = 2 (assuming 2 represents the save operation)

[0152] α E = 4, β E = 3, γ E = 2, δ E = 1

[0153] The calculation process is as follows:

[0154]

[0155] The calculated number 402365552 represents the unique code of the event. This number is converted into a form that is easy to store and query in the database through the combination of the timestamp, user ID, and behavior type.

[0156] Δt = 100 (assuming the server time is 100 seconds faster than the event time).

[0157] ∈ E = 10 (timestamp adjustment coefficient).

[0158] ζ U = 5, η U = 3 (user verification adjustment coefficient).

[0159] key = 2021 (the encryption key is assumed to be 2021).

[0160] Substitute into the formula and calculate:

[0161]

[0162] The timestamp 160945930 represents the adjusted time for the accuracy of the time in the log record. The user verification result 5036.67 is used to confirm the validity of the operation. This value is processed through the security key to verify the user ID.

[0163] Assume ξ L = 2, λ L = 1, combined with the previous calculation result Auth(U i ) ≈ 5036.67, substitute into the formula:

[0164]

[0165] The encrypted log entry value is 10073.34, which is used to ensure the security of log data, prevent unauthorized access, and finally upload it to the central database for storage and record, and output the event response record.

[0166] Please refer to Figure 4 , and the determination steps of the user behavior pattern are as follows:

[0167] Based on the event response record, extract the key user behavior data, including the access path and click frequency input data, and use the formula,

[0168]

[0169] Calculate the weighted value of user i in behavior j to generate the weighted user behavior data, where a ik represents the behavior data of user i in the kth activity of behavior j, m represents the total number of activities, and w k represents the weight coefficient of the kth activity;

[0170] Use the weighted user behavior data to standardize the data, and use the formula,

[0171]

[0172] Calculate the normalized score Z of the user behavior data ij , and generate the normalized user behavior data, where λ j is the normalization parameter of behavior j, μ j and σ j represent the average value and standard deviation of behavior j respectively;

[0173] Analyze the normalized user behavior data, and use the formula,

[0174]

[0175] Calculate the behavior pattern group Q assigned to user i to generate the user behavior pattern determination result, where ρ k is the adjustment coefficient of cluster k, G k represents the set of users assigned to the kth group, Z i represents the normalized behavior data vector of user i, and C k represents the center vector of the kth group, ∥Z i -C k ∥ represents the modulus length.

[0176] Assume that user i has three activities in behavior j, and the specific data a ij are [3, 5, 2] respectively, and the corresponding weights wk They are [0.5, 1, 0.5] respectively.

[0177] Substitute into the formula and calculate to get:

[0178] D ij = 0.5×3 + 1×5 + 0.5×2 = 1.5 + 5 + 1 = 7.5

[0179] The result D ij = 7.5 represents the total weighted behavior data of the user in behavior j, reflecting the comprehensive activity of the user in this behavior.

[0180] Assume that in behavior j, the average value μ j = 6, and the standard deviation σ j = 2, and the normalization adjustment coefficient λ j = 0.5.

[0181] Combined with the previous calculation result D ij = 7.5, substitute into the formula to get:

[0182]

[0183] The result Z ij = 1.5 represents the standardized score of user i in behavior j. This score takes into account the average level and variability of the behavior data, and more fairly reflects the relative performance of the user in behavior j.

[0184] Suppose there are two groups C 1 = [1.0, 0.5], C 2 = [2.0, 1.5], and the adjustment coefficient ρ 1 = 1, ρ 2 = 1.

[0185] Taking Z i = [1.5, 0.5] as an example, calculate the weighted distance from Z i to each C k :

[0186] C 1 = ∥[1.5, 0.5] - [1.0, 0.5]∥×1 = ∥0.5, 0∥ = 0.5

[0187] C 2 = ∥[1.5, 0.5] - [2.0, 1.5]∥×1 = ∥-0.5, -1.0∥ = 1.12

[0188] Then:

[0189] Q = min k (0.5, 1.12) = 0.5

[0190] User i is assigned to Group 1 because the distance to C 1 is relatively small.

[0191] Please refer to Figure 5 for the steps to obtain the record of document permission adjustment:

[0192] Based on the determination result of the user behavior pattern, set the access permission level for each behavior pattern group. Using the formula,

[0193]

[0194] calculate the access permission level P representing the k-th group k , and generate the group access permission configuration. Among them, V k is the behavior activity coefficient of group k, L k is the behavior center value of group k, reflecting the behavior pattern of the group, H k is the historical compliance score, θ k is the permission threshold of group k;

[0195] According to the group access permission configuration, update the document access permission database. Using the formula,

[0196]

[0197] calculate the permission update status value ZT and generate the real-time access permission update record. Among them, ∈ k is the sensitivity parameter, used to adjust the sensitivity of the update amplitude, and δ k is the update threshold, used to fine-tune the permission change;

[0198] Based on the real-time access permission update record, record the permission adjustment content of each user. Using the formula,

[0199]

[0200] calculate the ratio B of the user's last login time to the current time, and generate the document permission adjustment record. Among them, L u represents the last login time of user u, T is the current time, and τ is the time stamp of the permission update.

[0201] Assume that the parameters of group k are set as: V k = 0.8, L k = 1.2, H k = 0.3, θ k = 0.5.

[0202] Substitute into the formula and calculate to get:

[0203]

[0204] The result Pk = 1.2 indicates that the access permission level of group k is 1.2, which is a relatively high level, reflecting the active behavior and good compliance history of the group, and granting a relatively high access permission.

[0205] Combined with the foregoing calculation results, P k = 1.2, if δ k = 1.0, ∈ k = 0.1, substituting into the formula gives:

[0206]

[0207] The result ZT ≈ 0.182 indicates that the permission has been improved, and the updated status is positive, showing the actual improvement range of the permission.

[0208] Assume L u = 100 (unit: hours ago), T = 120 (unit: hours), τ = 110 (unit: hours), substituting into the formula, we get:

[0209]

[0210] The result B ≈ 0.917 represents the ratio from the user's last login to the current time. A value close to 1 indicates that the user is active, and the update operation occurred during the user's active period.

[0211] Please refer to Figure 6 , the analysis steps for the user behavior data trend are as follows:

[0212] Based on the document permission adjustment records, obtain the user ID, document ID, access time, and permission status, and use the formula,

[0213]

[0214] Calculate the total number of accesses D of user uid uid , generate user access count data, where v uid represents the number of accesses of user uid to document d, and n is the number of documents accessed;

[0215] Based on the user access count data, identify the user's access pattern, and use the formula,

[0216]

[0217] Calculate the weighted sum M of the average number of accesses per time unit of the user uid , determine the user's access frequency and preference, and generate the user access pattern analysis result, where D uid (t) is the number of accesses of the user at time t;

[0218] Based on the analysis results of the user access pattern, analyze the user behavior trend and use the formula,

[0219]

[0220] calculate the quantization value T of the change trend of user behavior uid , identify the changes and periodicity of the behavior pattern, and generate the analysis results of the user behavior trend. Among them, is the average value of M uid .

[0221] Suppose the user uid has accessed three documents, and the access times v uid are 2, 3, and 5 times respectively.

[0222] Then:

[0223] D uid = 2 + 3 + 5 = 10

[0224] The result D uid = 10 indicates that the user uid has accessed the document 10 times in total.

[0225] Suppose the user's access times in three different months are as follows:

[0226] 1 time in the first month, 3 times in the second month, and 2 times in the third month. Substitute into the formula to get:

[0227]

[0228] The result M uid ≈ 4.33 indicates that the weighted sum of the user's access times per time unit is 4.33, showing the distribution of access activities over time.

[0229] Continue to use the aforementioned calculation results. If Then:

[0230] T uid = (1 - 4.33) 2 + (6 - 4.33) 2 + (6 - 4.33) 2

[0231] = (-3.33) 2 + (1.67) 2 + (1.67) 2

[0232] ≈ 11.09 + 2.79 + 2.79 = 16.67

[0233] The result T uid=16.67 indicates that the user's behavior pattern shows obvious changes over time. A larger value indicates a larger magnitude of change, indicating the instability or development trend of user behavior.

[0234] See also Figure 7 , the steps to obtain document security management records are:

[0235] According to the analysis results of user behavior trends, the formula is used.

[0236] Δ uid =|T uid -S|

[0237] Calculate the deviation from normal behavior Δ uid , generating deviation detection results, where S is the preset standard behavior mode value;

[0238] Evaluate the deviation detection result to determine whether it exceeds the preset threshold Y. If Δ uid >Y, it is marked as abnormal, otherwise it is marked as normal and the abnormal behavior identification result is generated;

[0239] Based on the results of abnormal behavior identification, formulate and implement security measures and output document security management records.

[0240] Assume T uid is 16.67 (obtained from the previous calculation), and the standard mode value S is 10, then:

[0241] Δ uid =|16.67-10|=6.67

[0242] Results Δ uid =6.67 indicates that the user behavior deviates significantly from the standard pattern, indicating abnormal behavior or significant behavior changes.

[0243] Set the threshold value Y = 5, combined with the above calculation result Δ uid =6.67, we get:

[0244] Δ uid =6.67>Y=5

[0245] Marking user behavior as abnormal indicates the need to develop and implement security measures and output document security management records.

[0246] A digital flexible method for document management and control, comprising the following steps:

[0247] S1: Separate text content and structure, calculate word importance scores, identify core entities in documents, perform document monitoring, dynamically update and adjust the structure of the preliminary document dependency graph, and output the document semantic dependency graph;

[0248] S2: Based on the document semantic dependency graph, capture the operation events of the document, add the server timestamp, verify the user identity, perform encryption processing and upload it to the central database for storage and recording, and output the event response record;

[0249] S3: Based on the event response record, extract the key user behavior data, calculate the user behavior weighted value, perform normalization processing on the data, group the behavior patterns, and generate the user behavior pattern determination result;

[0250] S4: According to the user behavior pattern determination result, set the access permission level for each behavior pattern group, update the document access permission database, record the permission adjustment content of each user, and generate the document permission adjustment record;

[0251] S5: Based on the document permission adjustment record, calculate the total access times of the user, identify the access pattern of the user, determine the user access frequency and preference, analyze the user behavior trend, identify the changes and periodicity of the behavior pattern, and generate the user behavior trend analysis result;

[0252] S6: According to the user behavior trend analysis result, calculate the deviation value from the normal behavior, determine whether it exceeds the preset threshold and mark the user behavior, formulate and implement security measures, and output the document security control record.

[0253] The above is only the preferred embodiment of the present invention, and it does not limit the present invention in other forms. Any person skilled in the art may use the disclosed technical content to make changes or modifications into equivalent embodiments with equivalent changes and apply them to other fields. However, as long as it does not depart from the technical solution content of the present invention, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention still fall within the protection scope of the technical solution of the present invention.

Claims

1. A digital flexible system for document management and control, characterized by: The system comprises: The semantic construction module extracts keywords from documents, identifies key phrases, draws dependency graphs through connections between entities, monitors document updates, dynamically adjusts dependency relationships based on updates, reveals real-time connections between documents, and obtains document semantic dependency graphs; The event-driven module captures all operation events of the document based on the document semantic dependency graph, updates the event log with the captured data, responds to the event, records the event processing process, and generates event response records; The access control engine module performs statistical analysis based on user data in the event response records, determines user behavior patterns, updates access rights in real time based on the determination results, updates access logs synchronously, and obtains document permission adjustment records; The behavior analysis and response module analyzes the document permission adjustment records, extracts user access patterns, conducts user behavior data trend analysis, detects deviations from standard patterns based on the analysis results, identifies and responds to abnormal behaviors, formulates and implements security measures, and obtains document security management records; The steps for obtaining the document semantic dependency graph are: Separate text content and structure, remove irrelevant characters and stop words, and use formulas. Calculate the importance score TFIDF of word t imp (t, d), generate a set of keywords and key phrases, where ∈ is a small amount of fine-tuning the word frequency to avoid the zero frequency problem, δ is the adjustment amount for the total number of documents, d represents the document, N represents the total number of documents, DF(t) represents the number of documents including word t, and TF(t, d) is the word frequency of word t in document d; Based on the set of keywords and key phrases, the core entities in the document are identified using the formula, Calculate the dependency strength R between entities i and j adj (i, j), generate a preliminary document dependency graph, where α and β are regularization parameters used to balance the impact of entities on the results, CO(i, j) is the number of times entities i and j co-occur in the same document, and DN(i, j) is the number of documents that include entities i and j; Perform document monitoring, dynamically update and adjust the structure of the preliminary document dependency graph, using the formula, Calculate the importance score TFIDF of the changed word t dyn (t,d new ), generate a dynamically adjusted document dependency graph, where γ and η are adjustment parameters used to adjust the impact of document updates, d new represents the updated document, DC(t) represents the number of updated documents including word t, TF(t,d new ) is the updated document d new The frequency of word t in ; Based on the dynamically adjusted document dependency graph, the formula is adopted: Calculate the centrality Centr of document node v fin (v), and output the document semantic dependency graph, where ξ is the balance coefficient used to adjust the out-degree influence, v represents the document node, NB(v) represents other document nodes directly connected to the document node v, and DO(u) represents the number of documents directly connected to document u; The steps for determining the user behavior pattern are as follows: Based on the event response records, key user behavior data is extracted, including access path and click frequency input data, using the formula: Calculate the weighted value of user i in behavior j and generate user weighted behavior data, where a ik represents the behavior data of user i in the kth activity on behavior j, m represents the total number of activities, and w k represents the weight coefficient of the kth activity; Using the user weighted behavior data, the data is standardized and the formula is adopted: Calculate the normalized score Z of user behavior data ij , generate normalized user behavior data, where λ j is the normalization parameter of action j, μ j and σ j represent the mean and standard deviation of behavior j, respectively; The normalized user behavior data is analyzed using the formula: Calculate the behavior pattern group Q to which user i is assigned and generate the user behavior pattern determination result, where ρ k is the adjustment coefficient of cluster k, G k represents the set of users assigned to the kth group, Z i represents the normalized behavior data vector of user i, C k represents the center vector of the kth group, ∥Z i -C k ∥ indicates the module length; The steps for obtaining the document permission adjustment record are as follows: According to the user behavior pattern determination result, the access permission level is set for each behavior pattern group, using the formula: Calculate the access permission level P representing the kth group k , generate group access permission configuration, where V k is the behavioral activity coefficient of group k, L k is the behavior center value of group k, reflecting the behavior pattern of the group, H k is the historical compliance score, θ k is the permission threshold of group k; According to the group access permission configuration, the document access permission database is updated using the formula: Calculate the permission update state value ZT and generate real-time access permission update records, where ∈ k is a sensitivity parameter used to adjust the sensitivity of the update amplitude, δ k is the update threshold, used to fine-tune permission changes; Based on the real-time access authority update record, record the permission adjustment content of each user, using the formula, Calculate the ratio B of the user's last login to the current time and generate a document permission adjustment record, where L u represents the last login time of user u, T is the current time, and τ is the timestamp of the permission update.

2. The digital flexible system for document management and control according to claim 1 is characterized in that: The steps for obtaining the event response record are: Based on the document semantic dependency graph, the operation events of the document are captured, using the formula: Calculate the event code LogEntry(E i ), generate a preliminary event log, where α E , β E , γ E and δ E is an adjustment factor used to optimize the accuracy and flexibility of recording, E i represents the i-th event, T i represents the event time, U i Indicates user ID, A i Indicates the user operation type number; Based on the preliminary event log, the server timestamp is added and the user identity is verified using the formula, and Calculate the adjusted timestamp Timestamp(E i ) and the verified log entry value Auth(U i ), generating a timestamped and authenticated event log, where Δt represents the deviation between system time and event time, ∈ E , U and η U is the adjustment factor used to optimize the timestamp application and user verification accuracy, key is the encryption key; The timestamped and authenticated event log is encrypted using the formula, Calculate the encrypted log entry value Encrypt(AU i ), and get the encrypted event log, where AU i represents a verified log entry, ξ L and λ L is the coefficient used to adjust the encryption operation; Based on the encrypted event log, it is uploaded to the central database for storage and recording, and the event response record is output.

3. The digital flexible system for document management and control according to claim 1 is characterized in that: The analysis steps of the user behavior data trend are as follows: Based on the document permission adjustment record, the user ID, document ID, access time and permission status are obtained, and the formula is used: Calculate the total number of visits D by user uid uid , generate user access count data, where v uid represents the number of times user uid accesses document d, and n is the number of documents accessed; Based on the user access times data, the user's access pattern is identified, using the formula: Calculate the weighted sum of the average number of visits per time unit by the user M uid , determine the user's access frequency and preference, and generate the user access pattern analysis results, where D uid (t) is the number of visits by the user at time t; Based on the user access pattern analysis results, the user behavior trend is analyzed and the formula is used. Calculate the quantitative value T of the change trend of user behavior uid , identify the changes and periodicity of behavior patterns, and generate user behavior trend analysis results, where It is M uid The average value of .

4. The digital flexible system for document management and control according to claim 3 is characterized in that: The steps for obtaining the document security management record are as follows: According to the user behavior trend analysis results, the formula is used: Δ uid =|T uid -S| Calculate the deviation from normal behavior Δ uid , generating deviation detection results, where S is the preset standard behavior mode value; Evaluate the deviation detection result to determine whether it exceeds the preset threshold Y. If Δ uid >Y, it is marked as abnormal, otherwise it is marked as normal and the abnormal behavior identification result is generated; Based on the abnormal behavior identification results, safety measures are formulated and implemented, and document security management records are output.

5. A digital flexible method for document management, characterized in that: According to any one of claims 1 to 4, a digital flexible system for document management and control is implemented, comprising the following steps: Separate text content and structure, calculate word importance scores, identify core entities in documents, perform document monitoring, dynamically update and adjust the structure of the preliminary document dependency graph, and output the document semantic dependency graph; Based on the document semantic dependency graph, the document operation events are captured, a server timestamp is added, the user identity is verified, encryption is performed and uploaded to a central database for storage and recording, and event response records are output; Based on the event response record, extract key user behavior data, calculate user behavior weighted values, standardize the data, group behavior patterns, and generate user behavior pattern determination results; According to the user behavior pattern determination result, setting the access permission level for each behavior pattern group, updating the document access permission database, recording the permission adjustment content of each user, and generating the document permission adjustment record; Based on the document permission adjustment records, calculate the total number of user visits, identify the user's access pattern, determine the user's access frequency and preference, analyze user behavior trends, identify changes and periodicity of behavior patterns, and generate user behavior trend analysis results; Based on the user behavior trend analysis results, calculate the deviation value from the normal behavior, determine whether it exceeds the preset threshold and mark the user behavior, formulate and implement security measures, and output document security management records.

Images