Document collaboration method and system based on cloud office

Abstract

The invention provides a document collaboration method and system based on cloud office, and the method comprises the steps: firstly obtaining a target document operation record set of a cloud office platform, including user terminal operation event sequences (operation types and timestamps), and then recognizing document collaboration mode features according to the operation event sequences, operation overlapping time periods, type relevance and frequency distribution are covered, then a pre-training model is used for analysis to obtain a cooperation demand prediction result containing a cooperation object set and a cooperation strength index, dynamic permission configuration is conducted on cooperation objects according to the cooperation strength index, and a permission hierarchical structure is generated. And finally, in response to a cooperation request of the target user terminal, controlling the data synchronization process of the document based on the permission range. According to the method, the efficiency, the safety and the intelligent degree of cloud office document collaboration are effectively improved.

Description

Technical Field

[0001] The present application relates to the field of cloud office technology, and more specifically, to a document collaboration method and system based on cloud office. Background Art

[0002] With the rapid development of information technology, cloud office plays an increasingly important role in modern enterprises and team collaboration. In the cloud office environment, document collaboration is a core function.

[0003] Traditional cloud office document collaboration methods have many shortcomings. In early cloud office document collaboration, permission management mostly adopted a static and fixed mode. For example, administrators pre-set fixed permissions for users, such as editing, viewing, etc. This method lacks flexibility and cannot be dynamically adjusted according to the user's actual work needs and collaboration situation, which easily leads to unreasonable permission allocation. On the one hand, some users may be given too many unnecessary permissions, increasing the risk of documents being misoperated or maliciously tampered with; on the other hand, some users who really need more operating permissions have limited permissions, affecting work efficiency. Summary of the invention

[0004] In view of the above-mentioned problems, in combination with the first aspect of the present application, an embodiment of the present application provides a document collaboration method based on cloud office, the method comprising: Acquire a document operation record set generated by a target document in a cloud office platform, wherein the document operation record set includes a sequence of operation events performed by multiple user terminals on the target document, wherein the sequence of operation events includes at least one operation type and a corresponding operation timestamp; Based on the operation event sequence of each user terminal in the document operation record set, identifying the collaborative mode characteristics of the target document, the collaborative mode characteristics including the operation overlap period between multiple user terminals, the operation type correlation and the operation frequency distribution; The collaborative mode characteristics are analyzed by a pre-trained collaborative demand prediction model to generate a collaborative demand prediction result corresponding to the target document, wherein the collaborative demand prediction result includes a predicted collaborative object set and a collaborative strength index, and the collaborative strength index is used to characterize the degree of operation demand of each user terminal in the collaborative object set for the target document in a subsequent collaborative stage; Dynamically configuring permissions for user terminals in the collaboration object set according to the collaboration strength index, generating a permission hierarchy structure that matches the collaboration demand prediction result, wherein the permission hierarchy structure includes at least one permission level and a corresponding operation permission range, and the operation permission range is positively correlated with the collaboration strength index; In response to detecting a collaboration request of a target user terminal for the target document, based on the operation permission scope corresponding to the target user terminal in the permission hierarchy structure, a data synchronization process of the target document between the target user terminal and the cloud office platform is controlled.

[0005] On the other hand, an embodiment of the present application also provides a cloud office service system, including a processor and a machine-readable storage medium, wherein the machine-readable storage medium is connected to the processor, the machine-readable storage medium is used to store programs, instructions or codes, and the processor is used to execute the programs, instructions or codes in the machine-readable storage medium to implement the above method.

[0006] Based on the above aspects, the embodiment of the present application can comprehensively and accurately capture the operation of multiple user terminals on documents, including operation type and timestamp, by obtaining the document operation record set of the target document on the cloud office platform, so as to make the subsequent identification of collaboration mode features more accurate and detailed, and lay a solid foundation for accurate collaboration demand prediction. Compared with traditional document collaboration methods, it is no longer limited to simple operation record statistics, but deeply explores the pattern features behind the operation.

[0007] The collaborative pattern characteristics identified based on the operation event sequence cover multi-dimensional information such as operation overlapping time periods, operation type correlation, and operation frequency distribution, which helps to more deeply understand the user's behavior patterns in document collaboration, and overcomes the limitations of previous methods that can only evaluate the collaboration situation from a single dimension or a simple comprehensive dimension, thereby being able to more accurately adapt to complex and changeable cloud office document collaboration scenarios.

[0008] The pre-trained collaboration demand prediction model is used to analyze the characteristics of the collaboration mode and generate the collaboration demand prediction results. The predicted collaboration object set and collaboration intensity index provide forward-looking guidance for cloud office document collaboration, and can reasonably predict subsequent collaboration needs based on the user's previous operation mode in advance. This is a huge breakthrough in improving document collaboration efficiency. Traditional document collaboration methods can often only passively respond to immediate operation requests and lack the ability to effectively predict future collaboration needs, which can easily lead to problems such as unreasonable authority allocation and untimely data synchronization.

[0009] Dynamic permission configuration is performed based on the collaboration intensity index to generate a permission hierarchy structure, and the scope of permitted operations is positively correlated with the collaboration intensity index, so that permissions can be flexibly allocated based on the user's operational needs for the document, effectively avoiding the situation of over-allocation or insufficient permissions. While ensuring document security, it maximizes the operational flexibility of each user terminal in document collaboration, which is an advantage that traditional fixed permission allocation methods cannot match.

[0010] Finally, when responding to the collaboration request of the target user terminal, the data synchronization process is controlled based on the permission hierarchy structure. This precise data synchronization control method based on dynamic permissions ensures that each user terminal can efficiently interact with data within its permission range, reduces unnecessary data transmission and permission conflicts, improves the stability and efficiency of the entire cloud office document collaboration system, and further improves the overall performance of cloud office document collaboration. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 It is a schematic diagram of the execution flow of the cloud office-based document collaboration method provided in an embodiment of the present application.

[0012] Figure 2 It is a schematic diagram of the hardware architecture of the cloud office service system provided in an embodiment of the present application. DETAILED DESCRIPTION

[0013] The present application will be described in detail below with reference to the accompanying drawings. Figure 1 This is a flow chart of a document collaboration method based on cloud office provided by an embodiment of the present application. The document collaboration method based on cloud office is introduced in detail below.

[0014] Step S110, obtaining a document operation record set generated by the target document in the cloud office platform, wherein the document operation record set includes a sequence of operation events performed by multiple user terminals on the target document, and the operation event sequence includes at least one operation type and a corresponding operation timestamp.

[0015] In this embodiment, on a cloud office platform of a large enterprise, for example, there is a project planning document as a target document. This project involves multiple departments, including the marketing department, the R&D department, and the finance department. Employees of the marketing department use user terminal A, employees of the R&D department use user terminal B, and employees of the finance department use user terminal C. During the project advancement process, these user terminals will perform various operations on the target document.

[0016] User terminal A opens the target document at 9 am and views the content. The operation type of this operation is "document viewing" and the operation timestamp is "2023 - 05 - 01 09:00:00". Then at 10 am, the market research part of the document is edited. The operation type is changed to "document editing - market research part" and the operation timestamp is "2023 - 05 -01 10:00:00".

[0017] User terminal B also opened the target document for viewing at 10:30 a.m., with the operation type of "document viewing" and the operation timestamp of "2023 - 05 - 01 10:30:00". Then at 11 a.m., the technical solution part was edited, with the operation type of "document editing - technical solution part" and the operation timestamp of "2023 - 05 - 01 11:00:00".

[0018] User terminal C opens the target document at 1 pm to view budget-related content, the operation type is "document viewing-budget part", the operation timestamp is "2023-05-01 13:00:00", and modifies the budget data at 2 pm, the operation type is "document editing-budget part", and the operation timestamp is "2023-05-01 14:00:00".

[0019] The cloud office platform will record the sequence of operation events of these different user terminals on the target document, including the operation type and the corresponding operation timestamp, to form a document operation record collection. This collection comprehensively reflects the operations of employees from various departments on the target document at different times, providing basic data for subsequent analysis.

[0020] Step S120, based on the operation event sequence of each user terminal in the document operation record set, identifying the collaborative mode characteristics of the target document, the collaborative mode characteristics including the operation overlap period between multiple user terminals, the operation type correlation and the operation frequency distribution.

[0021] Continuing with the above-mentioned enterprise project planning document as an example, first extract the operation type sequence and the corresponding time interval sequence in the operation event sequence. For example, the operation type sequence of user terminal A is "document viewing-document editing-market research part", and its time interval sequence is 1 hour from viewing operation to editing operation. The operation type sequence of user terminal B is "document viewing-document editing-technical solution part", and the time interval is half an hour. The operation type sequence of user terminal C is "document viewing-budget part-document editing-budget part", and the time interval is 1 hour.

[0022] The operation overlap period is determined based on the distribution density of different types of operation events in the operation type sequence in the time dimension. In this project, between 10 a.m. and 10:30 a.m., user terminal A was editing the market research part, while user terminal B started to view the document. This is an operation overlap period. At this time, employees from different departments are interacting with the target document. Although the operation types are not exactly the same, they are all in the process of processing the document.

[0023] The context dependency between different operation types in the operation type sequence is analyzed to generate the operation type correlation. For example, after the marketing department completes the editing of the market research part (user terminal A), the technical solution editing of the R&D department (user terminal B) is often adjusted accordingly. After statistics of a large number of operation event sequences, it is found that after the editing of the market research part is completed, the probability of adjusting the technical solution editing within the next 2 hours is 60%, which reflects the correlation between operation types.

[0024] By counting the operation frequency distribution corresponding to different time differences in the time interval sequence, it is found that the operation frequency is higher from 9:00 to 11:00 every morning. During this period, the viewing and editing operations of the document are more frequent, while the operation frequency is lower after 3:00 in the afternoon. By combining these overlapping operation periods, the correlation of operation types, and the operation frequency distribution, the collaborative mode characteristics of the target document are obtained. These collaborative mode characteristics reflect the collaborative rules and mutual influence relationships of various departments in the project.

[0025] Step S130, analyzing the collaboration pattern characteristics through a pre-trained collaboration demand prediction model to generate a collaboration demand prediction result corresponding to the target document, wherein the collaboration demand prediction result includes a predicted collaboration object set and a collaboration strength index, and the collaboration strength index is used to characterize the degree of operation demand of each user terminal in the collaboration object set for the target document in a subsequent collaboration stage.

[0026] Still taking the enterprise's project planning document as an example, the previously identified collaboration pattern features, such as operation overlapping time periods, operation type correlation, and operation frequency distribution, are input into the pre-trained collaboration demand prediction model.

[0027] During the pre-training process, the model has learned a lot of historical collaboration data of similar project documents. For example, there were multiple project documents before, and each document had a corresponding set of final collaboration objects (such as which departments participated in the collaboration at each stage of the project) and actual collaboration intensity indicators (the degree of demand for document operations by each department at different stages). From these historical collaboration data, the historical operation records of each historical document are extracted, and a historical collaboration pattern feature set for training is generated. The initial neural network model is constructed and iteratively optimized to obtain this pre-trained collaboration demand prediction model.

[0028] When the collaboration mode features of the current project planning document are input into the feature coding layer of the model, a multi-dimensional feature vector is generated. For example, the duration coding of the overlapping operation period may reflect the overlapping operation duration of the marketing department and the R&D department from 10:00 to 10:30 in the morning, the probability coding of the operation type association reflects the probability that the marketing department's operation triggers the R&D department's operation, and the period coding of the operation frequency distribution indicates the periodic characteristics of frequent operations in the morning, etc.

[0029] The model's time series analysis layer models the time dependency of multi-dimensional feature vectors and outputs time series enhanced features. For example, the model predicts that within 2 hours after the marketing department completes the update of new market research results (9 am to 10 am) in the next week, the R&D department has an 80% probability of adjusting the technical solution, and the finance department may re-evaluate the budget between 1 pm and 2 pm on the same day.

[0030] The prediction layer is used to perform pattern matching on the time-series enhancement features to determine the historical and predicted collaboration contributions of each user terminal in the collaboration object set. Taking the marketing department as an example, its historical collaboration contribution is calculated based on the weight of the previous operation type in the document (such as the market research part has a higher editing weight) and the operation frequency, and the predicted collaboration contribution is calculated based on the collaboration event type associated with the marketing department in the time-series enhancement features (such as future market research updates) and the trigger time interval. The collaboration intensity index is generated based on the weighted sum of the historical collaboration contribution and the predicted collaboration contribution. For example, the marketing department has a higher collaboration intensity index, indicating that there is a greater demand for operations on the target document in the subsequent project promotion. The collaboration object set is sorted from high to low according to the collaboration intensity index to generate a collaboration demand prediction result. This collaboration demand prediction result can provide a basis for subsequent operations such as permission configuration.

[0031] Step S140, dynamically configure permissions for user terminals in the collaboration object set according to the collaboration strength index, and generate a permission hierarchy structure that matches the collaboration demand prediction result, wherein the permission hierarchy structure includes at least one permission level and a corresponding operation permission range, and the operation permission range is positively correlated with the collaboration strength index.

[0032] Based on the collaboration intensity index in the collaboration demand forecast results calculated earlier, dynamic permission configuration is performed for each department (user terminal) participating in the project planning document collaboration. For example, the marketing department is classified as a high collaboration intensity level due to its high collaboration intensity index. This level is configured with a wider range of first operation permissions, including document content editing permissions (the market research part and other related parts can be fully edited), version rollback permissions (if the previous edit is found to be incorrect, it can be rolled back to the previous version) and collaboration member invitation permissions (if you feel that other departments or personnel need to participate in the collaboration of specific content, you can invite them to join).

[0033] The R&D department's collaboration intensity index is at a medium level and is classified as a medium collaboration intensity level. The second operation permission configured includes document content editing permissions (but may be limited to the parts related to the technical solution) and annotation adding permissions (you can add annotations to the content in the document to explain your own ideas or opinions), and limits the frequency of use of version rollback permissions. For example, the version rollback operation can only be used once a month.

[0034] The collaboration intensity index of the Finance Department is relatively low, and is at a low collaboration intensity level. The configured third-operation permission scope only includes document content viewing permission (only the budget part and other related document contents can be viewed) and annotation viewing permission (annotations added by other departments can be viewed, but new annotations cannot be added).

[0035] The permission hierarchy structure is generated based on the mapping relationship between the permission level and the scope of operation permission. During the project progress, the changes in the collaboration intensity index are monitored in real time. For example, if the marketing department's demand for document operations suddenly decreases at a certain stage, the collaboration intensity index decreases, while the R&D department's operation demand increases and the collaboration intensity index increases, then the permission level allocation in the permission hierarchy structure will be dynamically adjusted, and some permissions of the marketing department may be adjusted to the R&D department to meet the actual collaboration needs of the project.

[0036] Step S150, in response to detecting a collaboration request of the target user terminal for the target document, based on the operation permission scope corresponding to the target user terminal in the permission hierarchy structure, controlling the data synchronization process of the target document between the target user terminal and the cloud office platform.

[0037] Assume that during the project, the marketing department staff uses the target user terminal A to initiate a collaboration request for the project planning document again. Since the marketing department is at a high collaboration intensity level, its operation permission range includes document content editing permissions. At this time, the cloud office platform will enable real-time synchronization mode. When the marketing department staff edits the target document on the target user terminal A, such as modifying the data in the market research section or adding new market analysis content, these editing operations will be uploaded to the cloud office platform in real time. In addition, because it is real-time synchronization, other online user terminals (such as user terminal B of the R&D department and user terminal C of the finance department) will receive content update notifications immediately and see the new edits of the marketing department.

[0038] If an employee of the Finance Department uses target user terminal C to initiate a collaboration request, the scope of operation permission does not include document content editing permissions due to its low collaboration intensity level. The cloud office platform will enable delayed synchronization mode and synchronize the read-only version of the target document to target user terminal C at a preset time interval (for example, every hour). In this way, employees of the Finance Department can only view the document content and cannot modify it. During the data synchronization process, the cloud office platform will record the operation logs of the target user terminal, such as the editing operation records of marketing department employees on target user terminal A or the viewing operation records of finance department employees on target user terminal C, and then merge these operation logs into the document operation record collection to update the collaboration mode features and provide new data basis for subsequent collaboration analysis and permission adjustment.

[0039] Based on the above steps, the embodiment of the present application can comprehensively and accurately capture the operation of multiple user terminals on documents, including operation type and timestamp information, by obtaining the document operation record set of the target document on the cloud office platform, so that the subsequent identification of collaboration mode features is more accurate and detailed, laying a solid foundation for accurate collaboration demand prediction. Compared with traditional document collaboration methods, it is no longer limited to simple operation record statistics, but deeply explores the pattern characteristics behind the operation.

[0040] The collaborative pattern characteristics identified based on the operation event sequence cover multi-dimensional information such as operation overlapping time periods, operation type correlation, and operation frequency distribution, which helps to more deeply understand the user's behavior patterns in document collaboration, and overcomes the limitations of previous methods that can only evaluate the collaboration situation from a single dimension or a simple comprehensive dimension, thereby being able to more accurately adapt to complex and changeable cloud office document collaboration scenarios.

[0041] The pre-trained collaboration demand prediction model is used to analyze the characteristics of the collaboration mode and generate the collaboration demand prediction results. The predicted collaboration object set and collaboration intensity index provide forward-looking guidance for cloud office document collaboration, and can reasonably predict subsequent collaboration needs based on the user's previous operation mode in advance. This is a huge breakthrough in improving document collaboration efficiency. Traditional document collaboration methods can often only passively respond to immediate operation requests and lack the ability to effectively predict future collaboration needs, which can easily lead to problems such as unreasonable authority allocation and untimely data synchronization.

[0042] Dynamic permission configuration is performed based on the collaboration intensity index to generate a permission hierarchy structure, and the scope of permitted operations is positively correlated with the collaboration intensity index, so that permissions can be flexibly allocated based on the user's operational needs for the document, effectively avoiding the situation of over-allocation or insufficient permissions. While ensuring document security, it maximizes the operational flexibility of each user terminal in document collaboration, which is an advantage that traditional fixed permission allocation methods cannot match.

[0043] Finally, when responding to the collaboration request of the target user terminal, the data synchronization process is controlled based on the permission hierarchy structure. This precise data synchronization control method based on dynamic permissions ensures that each user terminal can efficiently interact with data within its permission range, reduces unnecessary data transmission and permission conflicts, improves the stability and efficiency of the entire cloud office document collaboration system, and further improves the overall performance of cloud office document collaboration.

[0044] In a possible implementation, step S120 includes: Step S121: extracting an operation type sequence and a corresponding time interval sequence from the operation event sequence, wherein the time interval sequence represents a time difference between adjacent operation events.

[0045] In this embodiment, in the project planning document scenario of the enterprise, for user terminal A used by the marketing department, its operation event sequence is "document viewing" at 9 am and "document editing-market research part" at 10 am, then the operation type sequence is "document viewing, document editing-market research part", and the time interval sequence is 1 hour from 9 am to 10 am. The operation event sequence of user terminal B in the R&D department is "document viewing" at 10:30 am and "document editing-technical solution part" at 11 am, and the operation type sequence is "document viewing, document editing-technical solution part", and the time interval sequence is 0.5 hours from 10:30 am to 11 am. The operation event sequence of user terminal C in the finance department is "document viewing-budget part" at 1 pm and "document editing-budget part" at 2 pm, and the operation type sequence is "document viewing-budget part, document editing-budget part", and the time interval sequence is 1 hour. The extraction of these operation type sequences and time interval sequences provides basic data for subsequent analysis.

[0046] Step S122, determining the operation overlap period between the multiple user terminals based on the distribution density of different types of operation events in the operation type sequence in the time dimension, wherein the operation overlap period is a time interval in which at least two user terminals perform the same or related operation type on the target document within a preset time window.

[0047] For example, during the operation of this project planning document, from 10:00 to 10:30 a.m., user terminal A is editing the market research part, and user terminal B starts the document viewing operation. During this time period, user terminals from different departments interact with the target document. Although the operation types are not exactly the same, they are all in the process of processing related operations on the document, so this is an operation overlap period. For another example, from 1:00 to 2:00 p.m., the Finance Department is viewing and editing the budget part. If the Marketing Department also views the entire document again to check the relationship between the market research part and the budget, this is also an operation overlap period. These operation overlap periods reflect the collaborative attention of different departments to the document in the same time period, reflecting the potential collaborative relationship between departments.

[0048] Step S123, analyzing the context dependency between different operation types in the operation type sequence, generating the operation type relevance, wherein the operation type relevance is used to identify the probability that the first operation type triggers the second operation type, and the first operation type and the second operation type are executed by different user terminals.

[0049] For example, statistics from a large number of operation event sequences show that after the marketing department completes the editing of the market research part (user terminal A), the technical solution editing of the R&D department (user terminal B) will often be adjusted accordingly. After detailed data analysis, it is found that after the editing of the market research part is completed, the probability of adjusting the technical solution editing is 60% within the next 2 hours, which establishes the operation type correlation between the editing operation type of the marketing department and the editing operation type of the R&D department. Similarly, after the finance department completes the editing of the budget part, there is a 30% probability that the marketing department will review the entire document within the next 1 hour to adjust the correlation between the market research part and the budget, which also reflects the correlation between the operation types of different departments. This operation type correlation helps to deeply understand the mutual influence between the work of each department, so as to better grasp the collaborative process in the project.

[0050] Step S124: Count the operation frequency distributions corresponding to different time differences in the time interval sequence, and determine the active operation period and the low-activity period of the target document based on the operation frequency distributions.

[0051] For example, after statistically analyzing the time intervals of various operations during the entire project cycle, it was found that during the period from 9:00 a.m. to 11:00 a.m. every day, the frequency of operations on documents by various departments was relatively high. For example, the marketing department frequently checked and edited the market research part, the R&D department checked the document to prepare and edit the technical solution part, and the finance department checked the budget-related content. After 3:00 p.m., the frequency of operations decreased significantly, and the number of operations on documents by various departments decreased. 9:00 a.m. to 11:00 a.m. is the active operation cycle of the target document, and after 3:00 p.m. is the low-activity period. The determination of this active operation cycle and low-activity period helps to reasonably arrange resources and coordinate the work rhythm of various departments.

[0052] Step S125, combining the operation overlapping time period, operation type relevance and operation frequency distribution into the collaboration mode feature.

[0053] For example, in the collaborative mode characteristics of this project planning document, the overlapping operation period reflects the collaborative operation of different departments on the document in a specific time period, the operation type correlation reflects the causal or correlation relationship between the operation types of each department, and the operation frequency distribution clarifies the active and low-activity periods of document operations. These elements combined together comprehensively describe the collaborative mode characteristics of the project planning document. For example, the overlapping operation period shows the overlapping operations of the marketing department and the R&D department from 10 am to 10:30 am, the operation type correlation indicates the probability of the R&D department adjusting after the marketing department edits, and the operation frequency distribution determines that 9 am to 11 am is the active operation period. These combined fully depict the working mode and mutual relationship of each department in the project planning document collaboration, and provide an important basis for subsequent collaborative demand prediction, authority configuration and other operations.

[0054] In a possible implementation, step S130 includes: Step S131, input the collaboration mode feature into the feature coding layer of the collaboration demand prediction model to generate a multidimensional feature vector, which includes the duration coding of the operation overlap period, the probability coding of the operation type association and the period coding of the operation frequency distribution.

[0055] For the duration coding of the overlapping operation period, for example, the previously determined overlapping operation period between the marketing department and the R&D department from 10:00 to 10:30 am will be coded as a specific value, which reflects the length of the overlapping period and the relative position in the entire project cycle. The probability coding of the association of operation types, such as the probability that the R&D department will edit the technical solution part within 2 hours after the marketing department edits the market research part, is 60%. This probability will be encoded in a form suitable for model processing. The period coding of the operation frequency distribution, such as 9:00 to 11:00 am as the active operation period, will be encoded into a value containing period characteristics based on factors such as the operation frequency and operation type distribution within this period. These codes form a multi-dimensional feature vector.

[0056] Step S132, time-dependency modeling is performed on the multidimensional feature vector through the timing analysis layer of the collaboration demand prediction model, and timing enhancement features are output, wherein the timing enhancement features include the types of collaboration events that may be triggered by the target document within a preset time period in the future and the corresponding triggering time intervals.

[0057] In this embodiment, the timing analysis layer takes into account factors such as the order, time interval, and periodicity of operations. For example, based on the information in the multidimensional feature vector, the model may output that in the next week, when the marketing department completes the update of the new market research results between 9 and 10 a.m., the R&D department has an 80% probability of adjusting the technical solution within the next 2 hours, and the finance department may re-evaluate the budget between 1 and 2 p.m. on the same day. This information about the types of collaborative events that may be triggered by the target document in the future preset time period and the corresponding triggering time interval is the timing enhancement feature.

[0058] Step S133, use the prediction layer of the collaboration demand prediction model to perform pattern matching on the timing enhancement feature to determine the historical collaboration contribution and predicted collaboration contribution of each user terminal in the collaboration object set, the historical collaboration contribution is calculated based on the operation type weight and operation frequency of the user terminal in the historical operation event sequence, and the predicted collaboration contribution is calculated based on the collaboration event type and trigger time interval associated with the user terminal in the timing enhancement feature.

[0059] For example, for the marketing department, its historical collaboration contribution is calculated based on the operation type weight and operation frequency in the historical operation event sequence. For example, the editing operation weight of the market research part is relatively high, and the marketing department frequently performs editing operations in the market research part in the early stage of the project. Based on these operation type weights and operation frequencies, the historical collaboration contribution of the marketing department is calculated. The predicted collaboration contribution is calculated based on the collaboration event type and trigger time interval associated with the marketing department in the time series enhancement feature. As mentioned earlier, after the marketing department updates the market research results between 9 am and 10 am in the future, it may trigger related operations from the R&D department and the finance department. The predicted collaboration contribution of the marketing department is calculated based on these related operations and trigger time intervals.

[0060] Step S134, generating the collaboration intensity index according to the weighted sum of the historical collaboration contribution and the predicted collaboration contribution, and sorting the collaboration object set from high to low according to the collaboration intensity index to generate the collaboration demand prediction result.

[0061] For example, the historical collaboration contribution of the marketing department is 0.6, the predicted collaboration contribution is 0.4, and the weighted sum is 0.5 (assuming the weights are the same). This 0.5 is the collaboration intensity index of the marketing department. After calculating the collaboration intensity index of each user terminal (such as the R&D department, the finance department, etc.) in this way, sort the collaboration object set from high to low according to the collaboration intensity index to generate the collaboration demand prediction result. This collaboration demand prediction result can provide an important basis for subsequent operations such as permission configuration and resource allocation, which helps to more efficiently manage the collaboration process of project planning documents and improve the collaboration efficiency and quality of the entire project.

[0062] Wherein, the method further comprises: Step S101, obtaining historical collaboration data of multiple historical documents in the cloud office platform, wherein the historical collaboration data includes a final collaboration object set of each historical document and a corresponding actual collaboration intensity index.

[0063] For example, in this enterprise, there were multiple project documents before, such as past product promotion project documents, technical R&D project documents, etc. Each historical document has its final collaboration object set. For example, in the product promotion project document, the marketing department, sales department, and publicity department are the final collaboration object set; in the technical R&D project document, the R&D department and the testing department are the final collaboration object set. And each historical document has a corresponding actual collaboration intensity index. For the product promotion project document, the actual collaboration intensity index of the marketing department in the promotion stage is higher because they frequently conduct market research and formulate promotion strategies, while the actual collaboration intensity index of the sales department in the sales stage will increase, and the actual collaboration intensity index of the publicity department during the execution of the promotion activities reaches a peak. For the technical R&D project document, the actual collaboration intensity index of the R&D department in the development stage is the highest, and the actual collaboration intensity index of the testing department in the testing stage is higher.

[0064] Step S102: extracting the historical operation records of each historical document in the historical collaboration data, and generating a historical collaboration pattern feature set for training based on the historical operation records.

[0065] Taking the product promotion project document as an example, the historical operation records of the marketing department include the editing of the market research report section and the viewing of the promotion channel analysis section at different times; the historical operation records of the sales department include the editing of the sales data forecast section and the viewing of the customer feedback section; the historical operation records of the publicity department include the editing of the publicity copy section and the viewing of the publicity effect evaluation section. From the perspective of technical R&D project documents, the historical operation records of the R&D department involve the records of operations such as code writing and functional testing, and the historical operation records of the testing department mainly include the records of operations such as test case execution and defect reporting. Based on these historical operation records, a historical collaboration pattern feature set for training is generated. These historical collaboration pattern feature sets contain feature information such as operation overlap periods, operation type correlation, and operation frequency distribution in each historical document.

[0066] Step S103, constructing an initial neural network model, wherein the initial neural network model comprises a feature encoding layer, a time series analysis layer and a prediction layer.

[0067] Each layer of the initial neural network model has different functions. The feature encoding layer is responsible for encoding the input collaborative pattern features and converting them into a form suitable for subsequent analysis; the time series analysis layer focuses on modeling the time dependency of the data and mining the laws of the data in the time dimension; the prediction layer performs pattern matching and prediction operations based on the processing results of the previous two layers.

[0068] Step S104: input the historical collaboration pattern feature set into the initial neural network model, and iteratively optimize the model parameters of the initial neural network based on the final collaboration object set and the actual collaboration intensity index, until the error between the predicted collaboration intensity index output by the initial neural network model and the actual collaboration intensity index is lower than a preset threshold, thereby generating the pre-trained collaboration demand prediction model.

[0069] For example, for a product promotion project document, after inputting its historical collaboration pattern feature set into the model, the model parameters are adjusted according to the final collaboration object set of the marketing department, sales department, and publicity department and their respective actual collaboration intensity indicators. If the model initially predicts that the collaboration intensity indicator of the marketing department at a certain stage is significantly different from the actual value, the feature encoding layer will adjust the encoding method of the marketing department's operation type weight, the time series analysis layer will adjust the marketing department's operation time series, and the prediction layer will adjust the matching method of the marketing department's related patterns. Continue this process until the error between the predicted collaboration intensity indicator output by the initial neural network model and the actual collaboration intensity indicator is lower than the preset threshold, at which point a pre-trained collaboration demand prediction model is generated. This pre-trained model has learned the rules and patterns of collaboration of different project documents from a large amount of historical data.

[0070] In a possible implementation, step S123 includes: Step S1231: extracting a set of context windows of continuous operations from the operation type sequence, each context window comprising a target operation type and a preset number of adjacent operation types before and after it, wherein the adjacent operation types are executed by the same or different user terminals within a preset time difference range.

[0071] Take the operations of the marketing department, R&D department and finance department on the project planning document as an example. For the operation type sequence of the marketing department on user terminal A, such as "document viewing, document editing - market research part, document viewing - competitor analysis", assume that the preset number is 1 and the time difference range is 1 hour. Then a context window may be "document viewing - document editing - market research part", where "document editing - market research part" is the target operation type and "document viewing" is its previous adjacent operation type; another possible context window is "document editing - market research part - document viewing - competitor analysis", where "document editing - market research part" is the target operation type and "document viewing - competitor analysis" is its subsequent adjacent operation type. For the operation type sequence of the R&D department on user terminal B, "document viewing, document editing - technical solution part, document viewing - technical risk assessment", the context window is also extracted according to the preset rules. The operation type sequence of the finance department on user terminal C, "document viewing - budget part, document editing - budget part, document viewing - cost analysis", is also extracted in a similar way, so that a context window set containing the operations of each department is obtained.

[0072] Step S1232: Generate an operation co-occurrence matrix based on the co-occurrence frequency of the target operation type and the adjacent operation type in the context window set, wherein the rows of the operation co-occurrence matrix represent the first operation type performed by the first user terminal, and the columns represent the second operation type performed by the second user terminal, and the matrix element values ​​represent the co-occurrence frequency distribution of the first operation type and the second operation type in the context window set.

[0073] For example, the marketing department performs "document editing - market research part" (as the first operation type), and the R&D department performs "document editing - technical solution part" (as the second operation type), and their co-occurrence frequencies are counted in many context windows. If there are 30 times in 100 context windows where these two operation types appear in their respective operation sequences at the same time, then the corresponding matrix element value in the operation co-occurrence matrix (i.e., co-occurrence frequency distribution) is 30. Assuming that the operation types of the marketing department are rows and the operation types of the R&D department are columns, the matrix elements are counted and filled in turn to construct a complete operation co-occurrence matrix. This operation co-occurrence matrix comprehensively reflects the co-occurrence of different operation types of different departments in the context window set.

[0074] Step S1233, identifying the conditional trigger relationship across user terminals based on the difference in co-occurrence frequency distribution in the row direction and the column direction in the operation co-occurrence matrix, wherein the conditional trigger relationship indicates the probability change amplitude of the second user terminal performing the second operation type within a subsequent preset time period when the first user terminal performs the first operation type.

[0075] For example, in the operation co-occurrence matrix, the co-occurrence frequency distribution of the row where the marketing department performs the operation type "document editing-market research part" is different from the co-occurrence frequency distribution of the column where the R&D department performs the operation type "document editing-technical solution part". Through detailed analysis, it is found that after the marketing department completes the "document editing-market research part", the frequency of the R&D department performing the "document editing-technical solution part" changes significantly in the following 2 hours. Specifically, before the marketing department performs the "document editing-market research part", the frequency of the R&D department performing the "document editing-technical solution part" is relatively low, and within 2 hours after the marketing department completes the operation, the frequency of the R&D department performing the "document editing-technical solution part" increases significantly. The change in this frequency is the embodiment of the conditional trigger relationship across user terminals, which represents an impact relationship between the marketing department's operation and the R&D department's operation.

[0076] Step S1234, mark the operation type combination whose probability change amplitude in the conditional trigger relationship exceeds the preset change amplitude as a strongly associated operation pair, and calculate the time sensitivity coefficient of the strongly associated operation pair in the context window set, and the time sensitivity coefficient is used to characterize the time attenuation characteristics of the second operation type after the first operation type is executed.

[0077] Assuming that the preset change range is 30%, if the probability change range of the R&D department executing "Document Editing - Market Research Part" after the Marketing Department executes "Document Editing - Technical Solution Part" reaches 40%, then the two operation type combinations are marked as strongly associated operation pairs. For the operation type combinations marked as strongly associated operation pairs, calculate their time sensitivity coefficients in the context window set. For example, after the Marketing Department executes "Document Editing - Market Research Part", the time interval distribution of the R&D Department executing "Document Editing - Technical Solution Part" is: 60% of the times are executed within 1 hour, 30% of the times are executed within 2 hours, and 10% of the times are executed within 3 hours. As time goes by, the proportion of the R&D Department executing the corresponding operation gradually decreases, and this time decay characteristic is characterized by the time sensitivity coefficient. The calculated time sensitivity coefficient reflects the time dependency of the R&D Department executing "Document Editing - Technical Solution Part" after the Marketing Department executes "Document Editing - Market Research Part".

[0078] Step S1235, generating the operation type association according to the probability change amplitude and time sensitivity coefficient of the strongly associated operation pair, wherein the operation type association includes the conditional probability and effective triggering time interval of the first operation type triggering the second operation type, wherein the conditional probability is positively correlated with the probability change amplitude, and the effective triggering time interval is negatively correlated with the time sensitivity coefficient.

[0079] Taking the strongly associated operation pair of the marketing department and the R&D department as an example, the probability change range is 40%. According to the positive correlation between the conditional probability and the probability change range, it is concluded that the conditional probability of the first operation type ("document editing - market research part" of the marketing department) triggering the second operation type ("document editing - technical solution part" of the R&D department) is higher. At the same time, since the time sensitivity coefficient reflects the time decay characteristics of the R&D department after the operation of the marketing department, according to the negative correlation between the effective trigger time interval and the time sensitivity coefficient, it is concluded that the effective trigger time interval is shorter, for example, it may be 1.5 hours. In this way, the operation type correlation includes the conditional probability of the first operation type triggering the second operation type and the effective trigger time interval. This operation type correlation accurately depicts the dependency and triggering rules between the operation types of different departments, and provides an important basis for in-depth understanding of the collaboration mode of project planning documents. Through similar analysis of all relevant operation type combinations, the operation type correlation in the entire project planning document operation can be fully constructed, thereby providing strong support for subsequent collaborative demand prediction, permission configuration and other operations.

[0080] In a possible implementation, step S140 includes: Step S141: divide the user terminals in the collaboration object set into at least three authority levels according to the collaboration strength index, and the at least three authority levels include a high collaboration strength level, a medium collaboration strength level, and a low collaboration strength level.

[0081] For example, in the collaborative process of this project planning document, the marketing department has a high collaboration intensity index due to its frequent operations on market research, competition analysis and other contents in the early stage of the project, and the continuous update of the document content according to market changes in the future. It is classified as a high collaboration intensity level. The operations of the R&D department in the project are mainly concentrated on the formulation and adjustment of technical solutions. The operation frequency is slightly lower than that of the marketing department. The collaboration intensity index is at a medium level and is classified as a medium collaboration intensity level. The finance department mainly operates on budget and cost-related contents at specific stages. The frequency of operations and the degree of impact on the overall document are relatively small. The collaboration intensity index is low and it is classified as a low collaboration intensity level.

[0082] Step S142: configuring a first permission scope of operations for the high collaboration intensity level, wherein the first permission scope of operations includes document content editing authority, version rollback authority, and collaboration member invitation authority.

[0083] For example, for the marketing department at a high collaboration intensity level, the first scope of operation permission includes document content editing permissions, which enables the marketing department to comprehensively edit the market research section, competitor analysis section, and other sections related to the market strategy in the project planning document. For example, the marketing department can modify the market size estimation, market trend analysis, and other contents in the document based on the latest market research data. Version rollback permissions are also included, which is very useful when the marketing department finds that there are errors in the previous edits or needs to refer to the content of the previous version. For example, after modifying the market strategy section, the marketing department finds that it conflicts with the previous market research data, and can use the version rollback permission to return to the previous version and edit it again. The permission to invite collaborative members is also within this scope. If the marketing department feels that other departments (such as the public relations department) need to participate in the discussion of marketing strategies or the collaboration of document content, they can invite them to join.

[0084] Step S143, configuring a second permission range for the medium collaboration intensity level, wherein the second permission range includes document content editing authority and annotation adding authority, and limits the frequency of use of version rollback authority.

[0085] For example, the R&D department is at the medium collaboration intensity level, and its second operation permission scope includes document content editing authority, but it is mainly for editing the technical solution part. For example, the R&D department can edit the technical solution-related content such as the technical architecture and technology selection in the document according to the development of technology or project requirements. At the same time, the R&D department is also given the permission to add annotations, which helps the R&D department to add annotations to explain the relationship between the technical solution and market demand or to make technical suggestions when viewing other parts of the document (such as the market research part). However, the frequency of use of the version rollback permission is limited. It is assumed that the version rollback operation can only be used once a month. This is because the operation of the R&D department is relatively stable, unlike the marketing department, which needs to adjust frequently according to external changes. Limiting the frequency of use of the version rollback permission can ensure the relative stability of the document version.

[0086] Step S144: configuring a third permission range for the low collaboration intensity level, wherein the third permission range only includes document content viewing authority and annotation viewing authority.

[0087] For example, the Finance Department is at a low collaboration intensity level, and its third operation permission only includes the document content viewing permission and annotation viewing permission. This means that the Finance Department can only view the budget section, cost analysis section, and annotations added by other departments in the document. For example, the Finance Department can view the annotations on the marketing budget added by the Marketing Department in the market research section, or the annotations on the R&D cost estimate added by the R&D Department in the technical solution section, but cannot edit the document content or add new annotations.

[0088] Step S145, generating the permission hierarchy structure according to the mapping relationship between the permission level and the operation permission scope, and monitoring the change of the collaboration strength index in real time to dynamically adjust the permission level allocation in the permission hierarchy structure.

[0089] In this embodiment, this permission hierarchy structure clearly defines the scope of operation permissions of user terminals with different collaboration intensity levels in the project planning document. During the execution of the project, the changes in the collaboration intensity index will be monitored in real time to dynamically adjust the permission level allocation in the permission hierarchy structure. For example, as the project progresses, after the marketing department completes market research and strategy formulation, the demand for document operations gradually decreases, and the collaboration intensity index begins to decline. In the process of refining and optimizing the technical solution, the R&D department needs to refer more to the research results of the marketing department and make frequent adjustments to the technical solution, and the collaboration intensity index gradually increases.

[0090] Wherein, the method further comprises: Step S210: when it is detected that the collaboration mode feature of the target document is updated, the analysis process of the collaboration demand prediction model is re-executed to generate an updated collaboration demand prediction result.

[0091] For example, the marketing department conducted another large-scale market research in the middle of the project. The results of the research had a significant impact on the technical solution, budget and other aspects of the project, which led to the update of the collaboration mode characteristics. At this time, it is necessary to re-execute the analysis process of the collaboration demand prediction model to generate updated collaboration demand prediction results. By re-analyzing the collaboration mode characteristics, such as changes in factors such as operation overlap period, operation type relevance and operation frequency distribution, the collaboration demand prediction model will recalculate the collaboration intensity index of each user terminal.

[0092] Step S220, comparing the difference in the cooperation strength index before and after the update, if the change in the cooperation strength index of a user terminal exceeds a preset range, dynamic adjustment of the permission hierarchy structure is triggered.

[0093] Assuming the preset range is 30%, if the collaboration intensity index of the R&D department increases by 40% from the previous medium level, it exceeds the preset range. The permission level may be upgraded from the medium collaboration intensity level to the high collaboration intensity level by re-dividing the permission level according to the updated collaboration intensity index. Then send a permission change notification to the affected user terminal (here is the R&D department), which includes the changed operation permission scope (such as the current document content editing permission, version rollback permission, and collaboration member invitation permission) and the effective time (for example, it will take effect on the next working day after the notification is sent). This mechanism of dynamically adjusting the permission hierarchy structure can ensure that each user terminal has the appropriate operation permissions according to its actual collaboration needs at different stages of the project, thereby improving the project's collaboration efficiency and the effectiveness of document management.

[0094] Step S230: re-classify the authority levels according to the updated collaboration strength index, and send an authority change notification to the affected user terminals, wherein the authority change notification includes the changed operation permission scope and effective time.

[0095] Throughout the life cycle of a project, this dynamic permission configuration and adjustment of the permission hierarchy will be continuously optimized based on the actual progress of the project and the changing roles of various departments in document collaboration to adapt to the complexity and dynamism of the project, ensure smooth collaboration on project planning documents, and at the same time ensure the security and integrity of the documents, avoiding unnecessary operational risks and abuse of permissions.

[0096] In a possible implementation, step S150 includes: Step S151, determining the syncable data content and synchronization frequency of the target document according to the operation permission range currently held by the target user terminal.

[0097] Step S152: if the operation permission scope includes document content editing authority, the real-time synchronization mode is enabled, the editing operation of the target user terminal on the target document is uploaded to the cloud office platform in real time, and instant content update of other online user terminals is triggered.

[0098] Step S153: if the operation permission range does not include document content editing authority, a delayed synchronization mode is enabled to synchronize the read-only version of the target document to the target user terminal at a preset time interval, and restrict the modification operation of the document content.

[0099] Step S154, during the data synchronization process, recording the operation log of the target user terminal, and merging the operation log into the document operation record set to update the collaboration mode feature.

[0100] In this embodiment, for example, the target user terminal A used by the marketing department, because it is at a high collaboration intensity level, the scope of operation permission includes document content editing authority. For such a situation, the cloud office platform will determine that all the contents of the target document are syncable data content, because the marketing department has the right to edit any part of the document. The synchronization frequency is real-time update in real-time synchronization mode, because the marketing department plays an important role in the collaboration of project planning documents, and its editing operations need to be fed back to the cloud office platform in a timely manner and let other collaborators know. When the marketing department staff edits the target document on the target user terminal A, such as modifying a certain data in the market research part or updating the description of the market strategy, the cloud office platform will immediately obtain these editing operations and upload them in real time. At the same time, due to the use of real-time synchronization mode, other online user terminals, such as user terminal B of the R&D department and user terminal C of the finance department, will receive content update notifications immediately and see the new editing content of the marketing department. This ensures that each department can adjust its work content or provide feedback in a timely manner according to the latest information of the marketing department.

[0101] As for the target user terminal C used by the Finance Department, since it is at a low collaboration intensity level, the scope of permitted operations does not include document content editing permissions. The cloud office platform will determine that the syncable data content of the target document is a read-only version, that is, all document content except for editing operations. The synchronization frequency is performed at preset time intervals, such as once an hour. This means that the cloud office platform will synchronize the read-only version of the target document to the target user terminal C every hour. In addition, due to the limitations of its permitted operation scope, employees of the Finance Department cannot modify the document content on the target user terminal C and can only view the document content. In this mode, the Finance Department can obtain timely update information on documents from other departments without interfering with the stability of the documents.

[0102] During the data synchronization process, whether it is the real-time synchronization of the marketing department or the delayed synchronization of the finance department, the cloud office platform will record the operation log of the target user terminal. For the target user terminal A of the marketing department, the operation log will record each editing operation in detail, including the time of editing, the specific content of editing (such as modifying a data value in the market research part, or adding new market trend analysis content), and the operator information. For the target user terminal C of the finance department, the operation log will record the time of viewing the document, the part of the document viewed, and other information. Then, the cloud office platform will merge these operation logs into the document operation record collection to update the collaboration mode characteristics. For example, the frequent editing operations of the marketing department may affect the operation frequency distribution, further increasing its operation frequency within a certain period of time; the viewing operations of the finance department will also become part of the collaboration mode characteristics, reflecting the interaction between different departments and documents under different permissions.

[0103] In a possible implementation, the method further includes: Step S310: When it is detected that multiple user terminals simultaneously perform conflicting operations on the target document, conflict arbitration is performed based on the authority level of each user terminal in the authority hierarchy structure.

[0104] Step S320: preferentially retain the operation results of the user terminals with a high collaboration intensity level, and mark the conflicting operations as pending events.

[0105] Step S330: Send a conflict notification to all user terminals involved in the conflicting operation, where the conflict notification includes the conflicting operation type, conflicting content fragment, and suggested solution.

[0106] Step S340, in response to receiving a conflict resolution instruction submitted by any user terminal, adjusting the content of the target document according to the instruction, and updating the document operation record set and the collaboration mode characteristics.

[0107] During the project, multiple user terminals may perform conflicting operations on the target document at the same time. For example, an employee of the marketing department on target user terminal A and an employee of the R&D department on target user terminal B perform operations on a certain part of the project planning document almost at the same time. Assume that the marketing department is modifying the market size estimation data in the market research part, while the R&D department is adjusting the technical demand assessment part of the market size according to the technical solution at the same time, and these two operations conflict.

[0108] At this time, conflict arbitration is performed based on the permission level of each user terminal in the permission hierarchy. Since the marketing department is at a high collaboration intensity level and the R&D department is at a medium collaboration intensity level, the operation results of the user terminal with a high collaboration intensity level (i.e., the marketing department) are retained first according to the rules. The modification operation of the marketing department on the market size estimation data is regarded as a valid operation, while the operation of the R&D department is marked as a pending event.

[0109] Then, the cloud office platform will send a conflict notification to all user terminals involved in the conflicting operation. This conflict notification includes the type of conflicting operation (such as editing operations on market size-related content), conflicting content fragments (such as the market size estimate in the market research section and the related technical needs assessment section), and suggested solutions. For example, the suggested solution may be that the R&D department re-adjusts the technical needs assessment section based on the market size estimate revised by the marketing department, or that the marketing department and the R&D department negotiate to jointly determine a market size estimate and technical needs assessment content that both meets market demand and conforms to the technical solution.

[0110] When the cloud office platform receives a conflict resolution instruction submitted by any user terminal, it adjusts the content of the target document according to the instruction. If the R&D department readjusts the content of the technical requirements assessment part according to the modification of the marketing department, the cloud office platform will update the relevant content in the target document accordingly. In addition, in this process, the document operation record collection and collaboration mode characteristics will be updated. The document operation record collection will add records of this conflict operation, including information such as the time when the conflict occurred, the user terminals involved, and the conflict resolution process. The collaboration mode characteristics will also be updated according to these new operation records. For example, the operation type correlation may be affected, because this conflict operation and resolution process may reveal a new correlation between the operations of the marketing department and the R&D department; the operation frequency distribution may also change, especially the operation frequency in the time period involving the conflict operation will be adjusted. This mechanism for handling conflicting operations can ensure the accuracy and consistency of documents in the case of multi-department collaboration, and can also reasonably coordinate the working relationship between different departments to ensure the smooth progress of the project.

[0111] In a possible implementation, the method further includes: Step S410: generating a collaboration relationship map based on the collaboration demand prediction result, wherein the collaboration relationship map includes collaboration links, collaboration frequencies, and collaboration content relevance between user terminals in the collaboration object set.

[0112] In this embodiment, in the collaboration process of this project planning document, the collaboration demand prediction result includes information such as the predicted collaboration object set and the collaboration intensity index. Based on this information, a collaboration relationship map is constructed, which covers the collaboration links, collaboration frequencies and collaboration content relevance between user terminals in the collaboration object set. For example, the marketing department, the R&D department and the finance department are members of the collaboration object set, and there are different collaboration links between them. There is a collaboration link between the marketing department and the R&D department, because the market research results of the marketing department will affect the technical solution formulation of the R&D department, and the technical solution of the R&D department will also generate feedback on the market strategy adjustment of the marketing department. There is also a collaboration link between the marketing department and the finance department. The marketing promotion plan formulated by the marketing department involves budget and cost issues, which need to be determined in collaboration with the finance department, and the budget adjustment of the finance department will also affect the promotion strategy of the marketing department. There is also a collaboration link between the R&D department and the finance department, and the cost budget and funding allocation of the R&D project require the two to work together.

[0113] The frequency of collaboration can be determined by analyzing data such as document operation record collection. For example, the marketing department and the R&D department may interact several times a week. This interaction may be that the marketing department provides new market research data and the R&D department adjusts the technical plan accordingly, or the R&D department gives a technical feasibility assessment and the marketing department adjusts the market strategy, so the frequency of collaboration between them is relatively high. The finance department and the R&D department may only interact once or twice a month, mainly focusing on the adjustment of project budgets and the calculation of technical costs, and the frequency of collaboration is relatively low.

[0114] The relevance of collaborative content is reflected in the specific types of operations. The editing operations of the market research section of the marketing department are related to the editing operations of the technical solutions of the R&D department, because the market demand, competitive situation and other information in the market research are one of the bases for the R&D department to formulate technical solutions. The marketing promotion budget section of the marketing department is related to the budget editing operations of the finance department, and the technical cost estimation section of the R&D department is also related to the budget operations of the finance department. These collaborative links, collaboration frequencies and collaboration content relevance together constitute a collaborative relationship map, which comprehensively shows the relationship between departments in the project planning document collaboration.

[0115] Step S420, analyze the high-frequency collaborative links and key user nodes in the collaborative relationship graph, and generate collaborative optimization suggestions, wherein the collaborative optimization suggestions include increasing the authority level of the specified user terminal, splitting overloaded collaborative links, or merging low-frequency collaborative links, wherein the high-frequency collaborative link is a collaborative link with a collaborative frequency greater than a first set frequency, and the low-frequency collaborative link is a collaborative link with a collaborative frequency less than a second set frequency, and the first set frequency is greater than the second set frequency.

[0116] Assume that the first set frequency is set to 5 times per week, and the collaboration links with a frequency higher than this are considered high-frequency collaboration links. The collaboration link between the marketing department and the R&D department may be a high-frequency collaboration link because the number of interactions per week exceeds 5 times. Key user nodes are departments with significant influence in the collaboration relationship map. For example, the marketing department and the R&D department play a key leading and supporting role in the entire project, and they are key user nodes.

[0117] Some potential optimization points can be found by analyzing high-frequency collaborative links and key user nodes. For example, because the collaboration between the marketing department and the R&D department is very frequent and important, information overload or low collaboration efficiency may occur. At this time, collaborative optimization suggestions may include increasing the authority level of designated user terminals. If the R&D department needs to obtain the latest research information from the marketing department more promptly during the formulation of the technical solution and needs to adjust the content of related documents more independently, then consider increasing the authority level of the R&D department, such as granting more document content editing permissions or relaxing the restrictions on version rollback permissions, so that the R&D department can better collaborate with the marketing department.

[0118] For overloaded collaboration links, such as frequent information exchange between the marketing department and the R&D department, which leads to too frequent document updates, it may affect the needs of other departments to obtain stable information. At this time, the collaboration optimization suggestion may be to split the overloaded collaboration link. For example, the collaboration between the marketing department and the R&D department can be split according to different project stages or content modules, and the collaboration can be carried out separately at specific stages or for specific content modules, reducing unnecessary real-time interactions and improving overall collaboration efficiency.

[0119] For low-frequency collaboration links, assume that the second set frequency is set to once a month, and collaboration links with a frequency lower than this are considered low-frequency collaboration links. For example, the collaboration link between the Finance Department and the R&D Department may be a low-frequency collaboration link. In this case, you can consider merging low-frequency collaboration links. For example, you can merge the sporadic interactions between the Finance Department and the R&D Department on budgets and costs into a fixed time period or process to reduce unnecessary communication costs and waste of resources.

[0120] Step S430: Push the collaboration optimization suggestion to the administrator terminal of the cloud office platform, and execute the corresponding collaboration link adjustment operation in response to the administrator's confirmation instruction.

[0121] In this embodiment, after receiving these collaboration optimization suggestions, the administrator can perform macro control and adjustment on the collaboration process and permission settings of the project. For example, after seeing the suggestion to increase the permission level of the R&D department, the administrator can evaluate the overall situation of the project and consider whether it is really necessary to give the R&D department more permissions to improve the collaboration efficiency between the marketing department and the R&D department. If the administrator confirms that the suggestion is reasonable, a confirmation instruction will be issued.

[0122] If the administrator confirms to increase the permission level of the R&D department, the cloud office platform will adjust the scope of operation permission for the R&D department in accordance with the regulations, such as adding document content editing permissions or relaxing restrictions on version rollback permissions. If the collaboration link between the marketing department and the R&D department is split, the cloud office platform will re-plan the collaboration process between the two at different stages or modules to ensure the orderly transmission of information and the stable update of documents. If the low-frequency collaboration link between the finance department and the R&D department is merged, the cloud office platform will set up a special process and time period to centrally handle budget and cost-related matters between the two.

[0123] In a possible implementation, the method further includes: Step S510, embedding a collaborative visualization component in the user interface of the cloud office platform, wherein the collaborative visualization component dynamically displays the collaborative demand prediction result, the authority hierarchy structure, and the real-time data synchronization status.

[0124] For example, for the collaborative demand prediction results, the visualization component can display the collaborative intensity indicators of each user terminal in the form of charts or graphs, and intuitively show the degree of collaborative demand of the marketing department, R&D department, and finance department in the project. In terms of the permission hierarchy structure, the visualization component can use different colors or icons to represent different permission levels. For example, the marketing department with a high collaborative intensity level is marked in red, indicating that it has a wide range of operation permissions, including document content editing permissions, version rollback permissions, and collaborative member invitation permissions; the R&D department with a medium collaborative intensity level is marked in yellow, indicating that it has document content editing permissions (partial restrictions), annotation adding permissions, and limits the frequency of version rollback permissions; the finance department with a low collaborative intensity level is marked in green, indicating that it only has document content viewing permissions and annotation viewing permissions.

[0125] The real-time data synchronization status can also be displayed in the visualization component. For example, when the marketing department edits a document on user terminal A and synchronizes it to the cloud office platform in real time, the visualization component can display a dynamic arrow to indicate that the data flows from the marketing department's user terminal to the cloud office platform, and the real-time updated prompt information can be seen on other online user terminals (such as R&D department's user terminal B and finance department's user terminal C).

[0126] Step S520: providing a trend analysis diagram of the collaboration intensity index, a permission level distribution diagram, and an operation conflict hotspot diagram through the collaboration visualization component.

[0127] Step S530 , in response to a screening operation triggered by the user through the collaboration visualization component, filtering and displaying user terminals and their operation records in a specified authority level or collaboration intensity range.

[0128] The collaboration visualization component can also provide a trend analysis chart of the collaboration intensity index. This trend analysis chart can show the changes in the collaboration intensity index of each user terminal at different stages of the project. For example, in the early stage of the project, the collaboration intensity index of the marketing department gradually increased because a lot of market research and strategy formulation work was required; as the project progresses to the technical research and development stage, the collaboration intensity index of the R&D department begins to rise, while the collaboration intensity index of the marketing department may decrease slightly. The permission level distribution chart displays the permission level distribution of each user terminal in an intuitive way, so that all people involved in the collaboration can clearly understand the scope of their own and other departments' permissions. The operation conflict hotspot map can display the parts or time periods that are prone to conflicts during the document operation process. For example, if the marketing department and the R&D department often have conflicts in the market research part and the technical solution-related part, the operation conflict hotspot map will mark these two parts or related operation time periods to remind relevant personnel to pay attention.

[0129] When a user triggers a filtering operation through the collaborative visualization component, the user terminals and their operation records of the specified permission level or collaboration intensity range can be filtered and displayed. For example, if the administrator wants to view the operation records of the user terminals (R&D department) with a medium collaboration intensity level, the filtering condition can be set to the medium collaboration intensity level on the visualization component, and then the visualization component will only display the operation records of the R&D department, including the R&D department's editing operations on documents at different times, annotation addition operations, and interactions with other departments. This filtering function helps users quickly obtain relevant information according to their needs, and improve their understanding of project collaboration and management efficiency. Through these functions, the collaborative visualization component provides a comprehensive, intuitive, and convenient tool for project collaboration management, which helps to improve the collaboration efficiency and management level of the entire project.

[0130] Figure 2 The hardware structure of the cloud office service system 100 provided in the embodiment of the present application for implementing the above-mentioned cloud office-based document collaboration method is shown. Figure 2 As shown, the cloud office service system 100 may include a processor 110 , a machine-readable storage medium 120 , a bus 130 , and a communication unit 140 .

[0131] In one possible design, the cloud office service system 100 can be a single server or a server group. The server group can be centralized or distributed (for example, the cloud office service system 100 can be a distributed system). In some embodiments, the cloud office service system 100 can be local or remote. For example, the cloud office service system 100 can access information and / or data stored in a machine-readable storage medium 120 via a network. For another example, the cloud office service system 100 can be directly connected to the machine-readable storage medium 120 to access the stored information and / or data. In some embodiments, the cloud office service system 100 can be implemented on a server. By way of example only, the server may include a private cloud, a public cloud, a hybrid cloud, a community cloud, a distributed cloud, an internal cloud, a multi-layer cloud, etc., or any combination thereof.

[0132] The machine-readable storage medium 120 may store data and / or instructions. In some embodiments, the machine-readable storage medium 120 may store data obtained from an external terminal. In some embodiments, the machine-readable storage medium 120 may store data and / or instructions used by the cloud office service system 100 to execute or use to complete the exemplary methods described in this application.

[0133] During the specific implementation process, one or more processors 110 execute computer executable instructions stored in the machine-readable storage medium 120, so that the processor 110 can execute the document collaboration method based on cloud office in the above method embodiment. The processor 110, the machine-readable storage medium 120 and the communication unit 140 are connected through the bus 130, and the processor 110 can be used to control the sending and receiving actions of the communication unit 140.

[0134] The specific implementation process of the processor 110 can refer to the various method embodiments executed by the above-mentioned cloud office service system 100. The implementation principles and technical effects are similar, and this embodiment will not be repeated here.

[0135] In addition, an embodiment of the present application also provides a readable storage medium, in which computer executable instructions are preset. When the processor executes the computer executable instructions, the above-mentioned cloud office-based document collaboration method is implemented.

[0136] It should be noted that in order to simplify the description disclosed in the present application and thereby facilitate the understanding of one or more embodiments of the invention, in the foregoing description of the embodiments of the present application, various features are sometimes grouped into one embodiment, drawing or description thereof.

Claims

1. A document collaboration method based on cloud office, characterized in that: The method comprises: Acquire a document operation record set generated by a target document in a cloud office platform, wherein the document operation record set includes a sequence of operation events performed by multiple user terminals on the target document, wherein the sequence of operation events includes at least one operation type and a corresponding operation timestamp; Based on the operation event sequence of each user terminal in the document operation record set, identifying the collaborative mode characteristics of the target document, the collaborative mode characteristics including the operation overlap period between multiple user terminals, the operation type correlation and the operation frequency distribution; The collaborative mode characteristics are analyzed by a pre-trained collaborative demand prediction model to generate a collaborative demand prediction result corresponding to the target document, wherein the collaborative demand prediction result includes a predicted collaborative object set and a collaborative strength index, and the collaborative strength index is used to characterize the degree of operation demand of each user terminal in the collaborative object set for the target document in a subsequent collaborative stage; Dynamically configuring permissions for user terminals in the collaboration object set according to the collaboration strength index, generating a permission hierarchy structure that matches the collaboration demand prediction result, wherein the permission hierarchy structure includes at least one permission level and a corresponding operation permission range, and the operation permission range is positively correlated with the collaboration strength index; In response to detecting a collaboration request of a target user terminal for the target document, based on the operation permission scope corresponding to the target user terminal in the permission hierarchy structure, a data synchronization process of the target document between the target user terminal and the cloud office platform is controlled.

2. The document collaboration method based on cloud office according to claim 1, characterized in that: The identifying the collaborative mode feature of the target document based on the operation event sequence of each user terminal in the document operation record set includes: Extracting an operation type sequence and a corresponding time interval sequence from the operation event sequence, wherein the time interval sequence represents a time difference between adjacent operation events; Determining, according to the distribution density of different types of operation events in the operation type sequence in the time dimension, an operation overlap period between the multiple user terminals, wherein the operation overlap period is a time interval in which at least two user terminals perform the same or related operation type on the target document within a preset time window; Analyzing the context dependency between different operation types in the operation type sequence to generate the operation type relevance, where the operation type relevance is used to identify a probability that a first operation type triggers a second operation type, where the first operation type and the second operation type are executed by different user terminals; Counting the operation frequency distribution corresponding to different time differences in the time interval sequence, and determining the active operation period and the low activity period of the target document based on the operation frequency distribution; The operation overlapping period, operation type association and operation frequency distribution are combined into the collaboration mode feature.

3. The document collaboration method based on cloud office according to claim 2, characterized in that: The analyzing the collaboration mode characteristics by the pre-trained collaboration demand prediction model to generate a collaboration demand prediction result corresponding to the target document includes: Inputting the collaboration mode feature into the feature coding layer of the collaboration demand prediction model to generate a multidimensional feature vector, wherein the multidimensional feature vector includes the duration coding of the operation overlap period, the probability coding of the operation type association, and the period coding of the operation frequency distribution; The multidimensional feature vector is modeled for time dependency through the timing analysis layer of the collaboration demand prediction model, and a timing enhancement feature is output, wherein the timing enhancement feature includes the type of collaboration event that may be triggered by the target document within a preset time period in the future and the corresponding triggering time interval; Using the prediction layer of the collaboration demand prediction model to perform pattern matching on the timing enhancement feature, determine the historical collaboration contribution and predicted collaboration contribution of each user terminal in the collaboration object set, the historical collaboration contribution is calculated based on the operation type weight and operation frequency of the user terminal in the historical operation event sequence, and the predicted collaboration contribution is calculated based on the collaboration event type and triggering time interval associated with the user terminal in the timing enhancement feature; Generating the collaboration intensity index according to the weighted sum of the historical collaboration contribution and the predicted collaboration contribution, and sorting the collaboration object set from high to low according to the collaboration intensity index to generate the collaboration demand prediction result; Wherein, the method further comprises: Acquire historical collaboration data of multiple historical documents in the cloud office platform, wherein the historical collaboration data includes a final collaboration object set of each historical document and a corresponding actual collaboration intensity index; Extracting the historical operation record of each historical document in the historical collaboration data, and generating a historical collaboration pattern feature set for training based on the historical operation record; Constructing an initial neural network model, wherein the initial neural network model comprises a feature encoding layer, a time series analysis layer and a prediction layer; The historical collaboration pattern feature set is input into the initial neural network model, and the model parameters of the initial neural network are iteratively optimized based on the final collaboration object set and the actual collaboration intensity index, until the error between the predicted collaboration intensity index output by the initial neural network model and the actual collaboration intensity index is lower than a preset threshold, thereby generating the pre-trained collaboration demand prediction model.

4. The document collaboration method based on cloud office according to claim 2, characterized in that: The step of analyzing the context dependency between different operation types in the operation type sequence to generate the operation type relevance includes: Extracting a context window set of continuous operations from the operation type sequence, each context window comprising a target operation type and a preset number of adjacent operation types before and after it, wherein the adjacent operation types are executed by the same or different user terminals within a preset time difference range; Based on the co-occurrence frequency of the target operation type and the adjacent operation type in the context window set, an operation co-occurrence matrix is ​​generated, wherein the rows of the operation co-occurrence matrix represent the first operation type performed by the first user terminal, the columns represent the second operation type performed by the second user terminal, and the matrix element values ​​represent the co-occurrence frequency distribution of the first operation type and the second operation type in the context window set; identifying, according to a difference in co-occurrence frequency distribution between row directions and column directions in the operation co-occurrence matrix, a conditional trigger relationship across user terminals, the conditional trigger relationship indicating a probability change amplitude of the second user terminal performing the second operation type within a subsequent preset time period when the first user terminal performs the first operation type; Marking a combination of operation types whose probability change amplitude in the conditional trigger relationship exceeds a preset change amplitude as a strongly associated operation pair, and calculating a time sensitivity coefficient of the strongly associated operation pair in the context window set, wherein the time sensitivity coefficient is used to characterize a time attenuation characteristic of the second operation type after the first operation type is executed; The operation type association is generated according to the probability change amplitude and time sensitivity coefficient of the strongly associated operation pair, and the operation type association includes the conditional probability and effective triggering time interval of the first operation type triggering the second operation type, wherein the conditional probability is positively correlated with the probability change amplitude, and the effective triggering time interval is negatively correlated with the time sensitivity coefficient.

5. The document collaboration method based on cloud office according to claim 1, characterized in that: The dynamically configuring the permissions of the user terminals in the collaboration object set according to the collaboration strength index to generate a permission hierarchy structure matching the collaboration demand prediction result includes: Dividing the user terminals in the collaboration object set into at least three authority levels according to the collaboration strength index, wherein the at least three authority levels include a high collaboration strength level, a medium collaboration strength level, and a low collaboration strength level; Configuring a first permission scope for the high collaboration intensity level, the first permission scope including document content editing authority, version rollback authority, and collaboration member invitation authority; Configuring a second permission range for the medium collaboration intensity level, the second permission range includes document content editing authority and annotation adding authority, and limiting the frequency of use of version rollback authority; Configuring a third permission range for the low collaboration intensity level, wherein the third permission range only includes a document content viewing permission and an annotation viewing permission; Generating the permission hierarchy structure according to the mapping relationship between the permission level and the operation permission scope, and monitoring the change of the collaboration strength indicator in real time to dynamically adjust the permission level allocation in the permission hierarchy structure; Wherein, the method further comprises: When it is detected that the collaboration mode feature of the target document is updated, the analysis process of the collaboration demand prediction model is re-executed to generate an updated collaboration demand prediction result; Comparing the difference between the collaboration strength indicators before and after the update, if the collaboration strength indicator of a user terminal changes by more than a preset range, triggering a dynamic adjustment of the permission hierarchy structure; The authority levels are reclassified according to the updated collaboration strength index, and an authority change notification is sent to the affected user terminals, wherein the authority change notification includes the changed operation permission scope and effective time.

6. The document collaboration method based on cloud office according to claim 1, characterized in that: The process of controlling data synchronization of the target document between the target user terminal and the cloud office platform includes: Determining the syncable data content and synchronization frequency of the target document according to the scope of the operation permission currently held by the target user terminal; If the operation permission scope includes document content editing authority, the real-time synchronization mode is enabled to upload the editing operation of the target user terminal on the target document to the cloud office platform in real time, and trigger instant content update of other online user terminals; If the operation permission range does not include the document content editing permission, the delayed synchronization mode is enabled to synchronize the read-only version of the target document to the target user terminal at a preset time interval, and restrict the modification operation of the document content; During the data synchronization process, the operation log of the target user terminal is recorded, and the operation log is merged into the document operation record set to update the collaboration mode feature.

7. The document collaboration method based on cloud office according to claim 6, characterized in that: The method further comprises: When it is detected that multiple user terminals simultaneously perform conflicting operations on the target document, conflict arbitration is performed based on the authority level of each user terminal in the authority hierarchy structure; Prioritize the retention of operation results of user terminals with high collaboration intensity levels, and mark conflicting operations as pending events; Sending a conflict notification to all user terminals involved in the conflicting operation, wherein the conflict notification includes the conflicting operation type, the conflicting content fragment, and a suggested solution; In response to receiving a conflict resolution instruction submitted by any user terminal, adjusting the content of the target document according to the instruction, and updating the document operation record set and the collaboration mode characteristics.

8. The document collaboration method based on cloud office according to claim 1, characterized in that: The method further comprises: Generate a collaboration relationship map based on the collaboration demand prediction result, wherein the collaboration relationship map includes collaboration links, collaboration frequencies, and collaboration content relevance between user terminals in the collaboration object set; Analyze the high-frequency collaborative links and key user nodes in the collaborative relationship graph to generate collaborative optimization suggestions, wherein the collaborative optimization suggestions include increasing the authority level of a designated user terminal, splitting an overloaded collaborative link, or merging a low-frequency collaborative link, wherein the high-frequency collaborative link is a collaborative link with a collaborative frequency greater than a first set frequency, and the low-frequency collaborative link is a collaborative link with a collaborative frequency less than a second set frequency, and the first set frequency is greater than the second set frequency; The collaboration optimization suggestion is pushed to the administrator terminal of the cloud office platform, and the corresponding collaboration link adjustment operation is performed in response to the administrator confirmation instruction.

9. The document collaboration method based on cloud office according to claim 1, characterized in that: The method further comprises: Embedding a collaborative visualization component in the user interface of the cloud office platform, wherein the collaborative visualization component dynamically displays the collaborative demand prediction results, the authority hierarchy structure, and the real-time data synchronization status; Providing trend analysis diagrams of collaboration intensity indicators, permission level distribution diagrams, and operation conflict hot spots diagrams through the collaboration visualization component; In response to a screening operation triggered by a user through the collaboration visualization component, user terminals and their operation records in a specified authority level or collaboration intensity range are filtered and displayed.

10. A cloud office service system, characterized in that: The cloud office service system includes a processor and a memory, the memory is connected to the processor, the memory is used to store programs, instructions or codes, and the processor is used to execute the programs, instructions or codes in the memory to implement the cloud office-based document collaboration method described in any one of claims 1 to 9 above.

Images