A task coordination management method and system based on an offline application

Abstract

The application discloses a kind of based on the task cooperation management method and system of offline application, it is related to offline data management technical field, the method includes: determining task cooperation type, and obtaining collaborative task resource package;Task cooperation system environment is generated based on collaborative task resource package;Receive collaborative data, the modification probability of collaborative data is predicted by the prediction model pre-trained in collaborative task resource package, and the collaborative data of high-risk prediction result is early warning and checking;According to the compression packing of collaborative data according to collaborative task type, and when restoring network connection, the packaged collaborative data is executed upload by priority scheduling mechanism.The fusion technical scheme of intelligent matching resource package, generating system environment with edge computing, pre-training model early warning and checking, priority scheduling upload effectively reduces resource package mismatch and parameter error rate, while reducing upload delay, improves offline cooperation efficiency and data reliability.

Description

Technical Field

[0001] This invention relates to the field of offline data management technology, and in particular to a task collaboration management method and system based on offline applications. Background Technology

[0002] In the field of project task collaboration management, traditional task collaboration systems rely on network environments for task initiation and data processing. However, in scenarios with no or weak network access, such as construction sites, multiple technical bottlenecks are exposed. In existing solutions, users need to manually match task types and resource packages. When faced with task descriptions containing technical terms, such as "quality inspection of bridge pile foundation concrete pouring," human error can easily lead to mismatches between data models and business rules. For example, mistakenly selecting a road construction resource package can cause quality inspection standards to become disconnected from bridge engineering requirements. Simultaneously, traditional offline systems can only passively record data and lack the ability to anticipate risks associated with parameter modifications. Construction personnel often rework data due to forgetting to modify key parameters or failing to notice the impact of parameter correlations. Industry statistics show that the incidence of such problems is approximately 19%, particularly prominent in multi-person collaboration scenarios. Furthermore, when the network recovers, traditional systems use direct upload methods. The simultaneous transmission of large amounts of offline data can easily cause network congestion, delaying the upload of critical data such as quality rectification orders. Moreover, the lack of optimized compression strategies for data types further exacerbates the network burden. In addition, the system environment generation relies on static template loading, which cannot be dynamically adjusted according to task semantics and device type. Furthermore, it lacks edge computing capabilities to verify the collected data in real time, requiring manual review after connecting to the network, which severely restricts the efficiency of offline collaboration. Summary of the Invention

[0003] In view of this, the present invention proposes a task collaboration management method and system based on offline applications, which can effectively solve the problems of inefficient resource matching and data errors caused by lack of data verification during data upload in offline scenarios. The present invention provides the following technical solution:

[0004] A task collaboration management method based on offline applications, the method comprising:

[0005] The task collaboration type is determined based on user information and task semantic parsing results, and the corresponding collaboration task resource package is obtained; the task collaboration system environment is initialized and generated based on the collaboration task resource package;

[0006] The system receives collaborative data uploaded by users through the task collaboration system environment, predicts the probability of modification of the collaborative data through the pre-trained prediction model in the collaborative task resource package, and issues warnings and performs verification for collaborative data with high-risk prediction results.

[0007] The collaborative data is compressed and packaged according to the type of collaborative task, and uploaded through a priority scheduling mechanism when the network connection is restored.

[0008] Optionally, determining the task collaboration type based on user information and task semantic parsing results, and obtaining the corresponding collaborative task resource package includes:

[0009] Obtain the task description input by the user, and perform semantic parsing on the task description using natural language processing technology to extract task keywords;

[0010] Based on the user permission level corresponding to the task keywords and user information, a preset task type mapping rule is matched to determine the corresponding task collaboration type.

[0011] Based on the task collaboration type, the corresponding collaboration task resource package is retrieved.

[0012] Optionally, the collaborative task resource package includes a data model preset file, a business model preset file, a collaborative form style preset file, and a collaborative task flow rule file.

[0013] Optionally, the initialization of the task collaboration system environment based on the collaborative task resource package includes:

[0014] Unzip the collaborative task resource package to obtain the data model preset file, business model preset file, and collaborative form style preset file;

[0015] The data structure of the task collaboration system is automatically created in the device's offline database based on the data model pre-set file. The data structure includes master-slave table relationships and initialization parameters.

[0016] Based on the business model pre-configured file, a business model object entity containing business rules is automatically generated in the offline application of the device. The business rules include numerical verification logic and process control rules.

[0017] The collaborative task form interface is automatically created in the offline application of the device according to the preset collaborative form style file. The form interface is associated with the data structure and business model object entity.

[0018] Load the edge computing module to complete the generation of the task collaboration system environment.

[0019] Optionally, the training method for the prediction model includes:

[0020] Extract historical collaborative data and corresponding modification records from the collaborative task resource package or offline database. The modification records include the historical collaborative data modification timestamp, the modification frequency of each parameter, and the modification records of related parameters.

[0021] A single-parameter modification frequency prediction model is trained using the LSTM time series algorithm, wherein the LSTM algorithm weights the historical modification frequencies using a time decay factor.

[0022] The FP-Growth association rule algorithm is used to train the association parameter modification records to generate a parameter association modification prediction model. The FP-Growth algorithm mines the association between parameters through confidence and improvement metric indicators, and determines whether there is a strong association between parameters through preset trigger conditions.

[0023] The single-parameter modification frequency prediction model and the parameter association modification prediction model are stored on a local device for real-time prediction in offline mode.

[0024] Optionally, the step of receiving collaborative data uploaded by users through the task collaboration system environment, predicting the modification probability of the collaborative data through a pre-trained prediction model in the collaborative task resource package, and issuing warnings and verifications for collaborative data with high-risk prediction results includes:

[0025] The modification probability value of the current collaborative data is calculated based on the single-parameter modification frequency prediction model, and the modification correlation between collaborative data is analyzed based on the parameter correlation modification prediction model.

[0026] When the modification probability value exceeds a preset threshold or when the associated parameter is detected to have been modified, a visual warning mark is made on the corresponding data field in the collaborative form interface.

[0027] The edge computing module in the task collaboration system environment is invoked to perform numerical logic verification or image feature recognition verification on the early warning data, generate verification results and associate them with the collaborative data.

[0028] Optionally, the step of compressing and packaging the collaborative data according to the collaborative task type, and uploading the packaged collaborative data through a priority scheduling mechanism when the network connection is restored includes:

[0029] Identify the type of the current collaborative task and obtain the corresponding compression strategy configuration file;

[0030] According to the compression strategy configuration file, the collaborative data is compressed, packaged, and encapsulated.

[0031] Determine the business priority level based on the type of collaborative task;

[0032] The modification probability value of the current collaborative data is obtained from the prediction model, and the prediction risk priority is calculated based on the confidence and lift of the prediction model related to the parameters.

[0033] A comprehensive priority index is generated based on the business priority level and the predicted risk priority. The encapsulated collaborative data is then arranged in descending order of the comprehensive priority index to form an upload queue.

[0034] Real-time monitoring of network quality status; dynamic adjustment of upload strategy based on network quality; use accelerated transmission protocol to upload collaborative data with high comprehensive priority index; cache collaborative data with medium and low comprehensive priority index to local queue and upload them in queue order.

[0035] The uploaded data is recorded using a blockchain hash value, and the data consistency is verified through a consensus algorithm after the network is connected.

[0036] This invention further discloses a task collaboration management system based on offline applications, comprising:

[0037] The resource package acquisition module is used to determine the task collaboration type based on user information and task semantic parsing results, and to acquire the corresponding collaborative task resource package.

[0038] The environment initialization module is used to initialize and generate the task collaboration system environment based on the collaborative task resource package;

[0039] The data early warning module is used to receive collaborative data uploaded by users through the task collaboration system environment, predict the probability of modification of the collaborative data through the pre-trained prediction model in the collaborative task resource package, and issue early warnings and verification for collaborative data with high risk prediction results.

[0040] The data upload module is used to compress and package the collaborative data according to the collaborative task type, and to upload the packaged collaborative data through a priority scheduling mechanism when the network connection is restored.

[0041] The present invention further discloses a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method.

[0042] The present invention further discloses an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the above-described method.

[0043] According to the technical solution of this invention, intelligent identification of task collaboration types and accurate acquisition of resource packages are achieved through semantic parsing and user permission matching. Compared with the traditional manual matching method, the error matching rate of resource packages can be significantly reduced. Based on the pre-built files within the resource package, data structures, business models, and form interfaces are automatically constructed. By loading an edge computing module, the system can perform real-time logical verification on uploaded collaborative data in offline mode, effectively reducing the problem of parameter upload errors. Furthermore, the prediction model trained by LSTM and FP-Growth algorithms can predict the probability of modification of current collaborative data and the impact of parameter correlation based on historical modification data. When a high-risk modification is detected, a visual warning is triggered, and the edge computing module is linked to perform a secondary verification, significantly reducing the omission of parameter modifications due to forgetfulness or negligence by construction personnel. Finally, a comprehensive priority index is generated based on task type and predicted risk, avoiding network congestion and critical data delays caused by traditional data upload methods. Attached Figure Description

[0044] For illustrative and not limiting purposes, the present invention will now be described in conjunction with embodiments and accompanying drawings, wherein:

[0045] Figure 1 This is a flowchart illustrating the task collaboration management method based on offline applications in an embodiment of the present invention.

[0046] Figure 2 This is a schematic diagram of the constituent modules of the offline application-based task collaboration management system in an embodiment of the present invention;

[0047] Figure 3 This is a schematic diagram of the structure of the electronic device in an embodiment of the present invention. Detailed Implementation

[0048] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present application.

[0049] It should be noted that, where there is no conflict, the embodiments and features of the embodiments in this application can be combined with each other. The embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0050] refer to Figure 1This embodiment discloses a task collaborative management method based on offline applications. This method is used to realize task data collaborative management in an offline state, and specifically includes the following steps:

[0051] S100: Determine the task collaboration type based on user information and task semantic parsing results, and obtain the corresponding collaboration task resource package. The collaboration task resource package enables the initiation, reporting, and processing of collaboration tasks in offline mode.

[0052] Specifically, when a user launches the task collaboration system on an offline device, they first select a collaborative task or enter a task description through the interactive interface. For example, in a construction site scenario, they might enter "Quality inspection of concrete pouring for K12+300 bridge pile foundations." The system uses an integrated Natural Language Processing (NLP) module to semantically analyze this description, extracting key terms such as "bridge pile foundation," "concrete pouring," and "quality inspection" using a word segmentation algorithm. It then uses a pre-set professional thesaurus to eliminate ambiguity, clarifying "pile foundation" as "bridge pile foundation" rather than "building pile foundation." Simultaneously, the system reads the current user's identity information, including at least their permission level. The extracted keywords and the user's permission level are then input into a pre-set task type mapping rule base. This rule base is built based on historical project data. For example, the task type "bridge concrete construction quality inspection" corresponds to "bridge pile foundation - concrete pouring - quality inspection - quality inspector permissions." After successfully matching the task type with the mapping rule base, the system loads the corresponding collaborative task resource package from the local cache based on the task type. This resource package is filtered according to permissions; for example, ordinary construction workers cannot access resource packages containing design drawings. This collaborative task resource package includes a data model pre-configured file, used to define the table structure of the example pile foundation testing data; a business model pre-configured file, used to embed the business logic of the example concrete strength calculation formula; a collaborative form style pre-configured file, used to adapt the inspection form template to the offline device interface; and a collaborative task flow rule file, used to specify the process that needs to be reviewed after the user fills in the form. Throughout the entire processing flow, the system uses semantic parsing contextual understanding capabilities to automatically associate similar terms (such as "pouring" matching "casting"), ensuring the accuracy of task type identification and avoiding resource package mismatch problems caused by traditional manual selection.

[0053] S200: Initialize and generate the task collaboration system environment based on the collaborative task resource package.

[0054] After obtaining the collaborative task resource package, it is first encrypted, decrypted, and its integrity verified. Once confirmed to be correct, the data model preset file, business model preset file, and collaborative form style file are extracted. Taking the "Bridge Pile Foundation Concrete Pouring Quality Inspection" task in step S100 as an example, the data model preset file contains the structural definitions of the pile foundation detection master table and the concrete strength slave table. The system automatically executes the table creation statements using offline database engines such as SQLite, creating a complete data structure containing master-slave foreign key relationships locally on the device, and initializing default fields according to preset parameters. The business model preset file encapsulates business rules such as the concrete strength conversion formula and age correction coefficient. The system automatically generates business object entities containing the above rules at the application layer through Java reflection, such as the "Concrete Strength Detection" class. This entity can call the CPU computing power of the edge computing module in real time to perform numerical verification to determine whether the measured strength meets the design requirements. The collaborative form style file uses JSON format to define interface elements. After system parsing, it adaptively generates the form interface based on the device type. For example, on a tablet, it displays fields such as pile foundation number and inspection point location in a grid layout, and integrates a photo control associated with the "defect image" field. JavaScript is used to bidirectionally bind form input items to data structure table fields and business object attributes. Furthermore, a lightweight edge computing module is loaded from the resource package. This module is automatically registered to the business model entity. For example, when a user takes a photo of a concrete surface, the edge computing module calls a pre-built crack recognition model in real time to classify the image, and the recognition result is automatically filled into the "defect type" field of the form. This ultimately forms a complete task collaboration system environment including data storage, business logic, and an interactive interface, supporting real-time data processing and rule validation even offline.

[0055] S300: Receive collaborative data uploaded by the user through the task collaboration system environment, predict the modification probability of the collaborative data through the pre-trained prediction model in the collaborative task resource package, and issue warnings and verifications for collaborative data with high-risk prediction results.

[0056] Users upload collaborative data through a collaborative form interface, and the system captures data input events in real time, triggering the invocation of the prediction model. Before invoking the prediction model, this implementation discloses a training method for the prediction model. Specifically, historical collaborative data and corresponding modification records are extracted from a collaborative task resource package or an offline database. These modification records include historical collaborative data modification timestamps, the modification frequency of each parameter, and related parameter modification records. The historical modification period of each parameter in the historical collaborative data is trained using the LSTM time series algorithm to generate a single-parameter modification frequency prediction model. The related parameter modification records are then trained using the FP-Growth association rule algorithm to uncover the modification correlations between parameters, generating a parameter association modification prediction model. The single-parameter modification frequency prediction model and the parameter association modification prediction model are stored on a local device for real-time prediction in offline conditions.

[0057] Specifically, a single-parameter modification frequency prediction model is trained using the LSTM time series algorithm. The LSTM algorithm weights historical modification frequencies using a time decay factor. The specific formula for weighting historical modification frequencies is as follows: Among them, P S (t) represents the modification probability of a single parameter at time t, α is the time decay factor used to control the weight decay of historical data, and t i f is the timestamp of the i-th parameter modification. i The frequency of the i-th parameter modification.

[0058] The FP-Growth association rule algorithm is used to train the association parameter modification records to generate a parameter association modification prediction model. The FP-Growth algorithm mines the association between parameters through confidence and improvement metric indicators, and determines whether there is a strong association between parameters through preset trigger conditions.

[0059] The confidence formula is as follows: This represents the probability that parameter B is modified simultaneously when parameter A is modified; the lift formula is: Used to represent the lifting degree of parameters A and B, to measure whether the correlation is significantly higher than the probability of independent modification; S AB S represents the frequency support of parameters A and B occurring simultaneously. A S represents the frequency support of parameter A occurring alone. B This refers to the frequency support of parameter B appearing alone. In this embodiment, the preset triggering condition is C. AB >θ and L AB >1, where θ is a preset confidence threshold. By setting C... AB >θ, to ensure that the confidence level of the association rule is high and to avoid accidental associations. This is achieved by setting L... AB>1 ensures that the correlation between A and B is significantly higher than in the random independent case, thereby reducing misjudgment. For example, in an engineering construction project, if parameters A (concrete strength) and B (age correction factor) satisfy C... AB =0.85 and L AB If the value is 1.2, the system will trigger an alert, prompting the user to modify B simultaneously to avoid data inconsistency caused by modifying A alone.

[0060] The system loads a trained LSTM single-parameter modification frequency model and an FP-Growth association model. The system inputs the currently submitted collaborative data into the LSTM model, calculates its modification probability through time-series pattern analysis, and simultaneously detects the association between the current collaborative data and other previously entered data using the FP-Growth model. When the modification probability exceeds a preset threshold or an anomaly in strongly correlated data is detected, the system issues a visual warning next to the corresponding data field in the form, along with a pop-up message: "This collaborative data has a high risk of modification; review is recommended." Subsequently, the system invokes the edge computing module within the task collaboration system environment to perform multi-dimensional verification of the warning data: verifying the data's logical consistency through the rule engine in the business model, and simultaneously enabling a lightweight AI model to perform feature recognition on image-based collaborative data. The verification results are then associated with the collaborative data record in the form of an encrypted timestamp, forming an intelligent processing flow covering prediction, warning, and verification, ensuring the accuracy and risk controllability of various types of collaborative data submitted offline.

[0061] S400: Compress and package the collaborative data according to the collaborative task type, and upload the packaged collaborative data through a priority scheduling mechanism when the network connection is restored.

[0062] Specifically, after identifying the current collaborative task type, the system retrieves the corresponding configuration file from a pre-set compression strategy library. Examples include: compressing image data in quality rectification tasks using WebP format and text data using the GZIP algorithm, forming a categorized compression strategy. Next, the system performs differentiated compression on the collaborative data according to this strategy and encapsulates it into a data packet containing a task type identifier and a compression timestamp. Simultaneously, it embeds a parameter modification risk marker generated by a prediction model into the packet body metadata. When the network connection is restored, the system first determines the business priority level based on the task type, and then generates a comprehensive priority index based on the modification probability value output by the prediction model. All data packets are then arranged in descending order of the comprehensive priority index to form an upload queue. In this embodiment, the business priority level is defined as P. biz And P biz ∈[1, 5]. For example, the priority level of a quality rectification task is level 5, and the priority level of a progress reporting task is level 3. The modification probability value P of the current collaborative data is obtained through the prediction model in step S300. S(t), and modify the confidence level C of the prediction model based on parameter correlation. AB and the degree of improvement L AB Calculate the predicted risk priority P risk The calculation formula is: Where τ is the preset modification probability threshold, and C max This represents the upper limit threshold for associated risk factors. The comprehensive priority index P is generated using a non-linear combination formula. total The calculation formula is: Where β∈[0,1], is the business priority weight coefficient. The encapsulated collaborative data is then sorted according to P... total The upload queue is formed by sorting the data in descending order.

[0063] Simultaneously, the system monitors network quality in real time. When network congestion is detected, an accelerated transmission protocol is activated for high-priority data packets. For example, segmented encrypted transmission is used to ensure that critical data with a high overall priority index is uploaded first, while medium- and low-priority data packets are cached in a local queue and automatically uploaded in queue order when the network becomes less congested. During the upload process, the system calculates a blockchain hash value for each data packet and records it locally. After reconnecting to the network, the system verifies data consistency with the cloud through a consensus algorithm, ensuring the integrity and reliability of data during transmission.

[0064] In summary, this implementation method achieves intelligent identification of task collaboration types and accurate acquisition of resource packages through semantic parsing and user permission matching, effectively reducing the error rate of resource packages under traditional manual matching methods and solving the problem of resource package mismatch caused by complex task descriptions. Based on pre-set resource package files, a task collaboration system environment containing data structures, business models, and form interfaces is automatically generated, and an edge computing module is integrated, enabling the system to perform real-time logical verification and AI feature recognition on collaborative data such as engineering inspection values ​​and images to ensure the accuracy of collaborative data. A closed-loop mechanism of "prediction-early warning-verification" built using LSTM and FP-Growth pre-trained models can predict data modification risks based on the temporal characteristics and association rules of historical collaborative data. When high-risk data is detected, a visual early warning is triggered, and the edge computing module is linked to perform dual verification, effectively reducing data errors caused by human error. During the data upload stage, a comprehensive priority index is generated based on the task type and predicted risk. Combined with network quality, the transmission strategy is dynamically adjusted, and accelerated transmission protocols are enabled for critical data, significantly reducing data upload latency. Simultaneously, classification compression avoids network congestion problems for uploading critical data with high comprehensive priority indices.

[0065] refer to Figure 2 This embodiment further discloses a task collaboration management system based on offline applications, including a resource package acquisition module 21, an environment initialization module 22, a data early warning module 23, and a data upload module 24. Detailed descriptions follow:

[0066] The resource package acquisition module 21 is used to determine the task collaboration type based on user information and task semantic parsing results, and to acquire the corresponding collaborative task resource package. This includes: acquiring the task description input by the user, and performing semantic parsing on the task description using natural language processing technology to extract task keywords; matching a preset task type mapping rule based on the task keywords and the user permission level corresponding to the user information to determine the corresponding task collaboration type; and retrieving the corresponding collaborative task resource package based on the task collaboration type.

[0067] The environment initialization module 22 is used to initialize and generate a task collaboration system environment based on the collaborative task resource package. This includes: decompressing the collaborative task resource package to obtain a data model preset file, a business model preset file, and a collaborative form style preset file; automatically creating a data structure for the task collaboration system in the device's offline database based on the data model preset file, the data structure including master-slave table relationships and initialization parameters; automatically generating a business model object entity containing business rules in the device's offline application based on the business model preset file, the business rules including numerical validation logic and flow control rules; automatically creating a form interface for the collaborative task in the device's offline application based on the collaborative form style preset file, the form interface being associated with the data structure and the business model object entity; and loading the edge computing module to complete the generation of the task collaboration system environment.

[0068] The data early warning module 23 is used to receive collaborative data uploaded by users through the task collaboration system environment, predict the modification probability of the collaborative data through a pre-trained prediction model in the collaborative task resource package, and issue early warnings and verifications for collaborative data with high prediction risks. This includes: calculating the modification probability value of the current collaborative data based on the single-parameter modification frequency prediction model; analyzing the modification correlation between collaborative data based on the parameter association modification prediction model; when the modification probability value exceeds a preset threshold or a modified associated parameter is detected, visually marking the corresponding data field in the collaborative form interface with an early warning; and calling the edge computing module in the task collaboration system environment to perform numerical logic verification or image feature recognition verification on the early warning data, generating verification results and associating them with the collaborative data.

[0069] The data upload module 24 is used to compress and package the collaborative data according to the collaborative task type, and upload the packaged collaborative data through a priority scheduling mechanism when the network connection is restored. This includes: identifying the type of the current collaborative task and obtaining the corresponding compression strategy configuration file; compressing and packaging the collaborative data according to the compression strategy configuration file; determining the business priority level according to the collaborative task type; obtaining the modification probability value of the current collaborative data from the prediction model, and generating a predicted risk priority based on the modification probability value; generating a comprehensive priority index based on the business priority level and the predicted risk priority, and arranging the packaged collaborative data in descending order of the comprehensive priority index to form an upload queue; monitoring the network quality status in real time, dynamically adjusting the upload strategy according to the network quality, using an accelerated transmission protocol to upload collaborative data with a high comprehensive priority index, caching collaborative data with a medium or low comprehensive priority index to a local queue and uploading it in queue order; recording the uploaded data using a blockchain hash value, and verifying data consistency through a consensus algorithm after network connection.

[0070] Figure 3 A schematic diagram of the physical structure of an electronic device provided in an embodiment of the present invention, such as... Figure 3 As shown, the electronic device 50 includes: a processor 501, a memory 502, and a bus 503;

[0071] The processor 501 and the memory 502 communicate with each other via the bus 503; the processor 501 is used to call the program instructions in the memory 502 to execute the methods provided in the above-described embodiments.

[0072] This embodiment provides a non-transitory computer-readable storage medium that stores computer instructions that cause a computer to execute the methods provided in the above-described embodiments.

[0073] Those skilled in the art will understand that all or part of the steps of the above-described method implementation can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it performs the steps of the above-described method implementation. The aforementioned storage medium includes various storage media capable of storing program code, such as ROM, RAM, magnetic disk, or optical disk.

[0074] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0075] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of each embodiment or some parts of the embodiments.

[0076] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can occur depending on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A task collaboration management method based on offline applications, characterized in that, The method includes: The task collaboration type is determined based on user information and task semantic parsing results, and the corresponding collaboration task resource package is obtained, including: decompressing the collaboration task resource package to obtain a data model preset file, a business model preset file, and a collaboration form style preset file; automatically creating the data structure of the task collaboration system in the device's offline database based on the data model preset file, the data structure including master-slave table relationships and initialization parameters; automatically generating a business model object entity containing business rules in the device's offline application based on the business model preset file, the business rules including numerical validation logic and flow control rules; automatically creating a form interface for the collaboration task in the device's offline application based on the collaboration form style preset file, the form interface being associated with the data structure and the business model object entity; and loading the edge computing module to complete the generation of the task collaboration system environment. The task collaboration system environment is initialized and generated based on the aforementioned collaborative task resource package; The system receives collaborative data uploaded by users within the task collaboration system environment. It then uses a pre-trained prediction model from the collaborative task resource package to predict the probability of modification of the collaborative data, and issues warnings and performs verification for collaborative data with high-risk predictions. The training method for the prediction model includes: extracting historical collaborative data and corresponding modification records from the collaborative task resource package or an offline database. These modification records include modification timestamps, modification frequencies of each parameter, and modification records of associated parameters. A single-parameter modification frequency prediction model is trained using an LSTM time series algorithm, where the LSTM algorithm weights the historical modification frequencies using a time decay factor. The associated parameter modification records are then trained using an FP-Growth association rule algorithm to generate a parameter association modification prediction model. The FP-Growth algorithm mines the correlation between parameters using confidence and improvement metrics, and determines whether a strong correlation exists between parameters using preset trigger conditions. The single-parameter modification frequency prediction model and the parameter association modification prediction model are stored on a local device for real-time prediction in offline conditions. The collaborative data is compressed and packaged according to the type of collaborative task, and uploaded through a priority scheduling mechanism when the network connection is restored.

2. The task collaborative management method based on offline applications according to claim 1, characterized in that, The step of determining the task collaboration type based on user information and task semantic parsing results, and obtaining the corresponding collaboration task resource package includes: Obtain the task description input by the user, and perform semantic parsing on the task description using natural language processing technology to extract task keywords; Based on the user permission level corresponding to the task keywords and user information, a preset task type mapping rule is matched to determine the corresponding task collaboration type. Based on the task collaboration type, the corresponding collaboration task resource package is retrieved.

3. The task collaborative management method based on offline applications according to claim 2, characterized in that, The collaborative task resource package includes a data model pre-set file, a business model pre-set file, a collaborative form style pre-set file, and a collaborative task flow rule file.

4. The task collaborative management method based on offline applications according to claim 1, characterized in that, The process of receiving collaborative data uploaded by users through the task collaboration system environment, predicting the probability of modification of the collaborative data using a pre-trained prediction model in the collaborative task resource package, and issuing warnings and verifications for collaborative data with high-risk prediction results includes: The modification probability value of the current collaborative data is calculated based on the single-parameter modification frequency prediction model, and the modification correlation between collaborative data is analyzed based on the parameter association modification prediction model. When the modification probability value exceeds a preset threshold or when the associated parameter is detected to have been modified, a visual warning mark is made on the corresponding data field in the collaborative form interface; The edge computing module in the task collaboration system environment is invoked to perform numerical logic verification or image feature recognition verification on the early warning data, generate verification results, and associate them with the collaborative data.

5. The task collaboration management method based on offline applications according to claim 1, characterized in that, The step of compressing and packaging the collaborative data according to the collaborative task type, and uploading the packaged collaborative data through a priority scheduling mechanism when the network connection is restored, includes: Identify the type of the current collaborative task and obtain the corresponding compression strategy configuration file; According to the compression strategy configuration file, the collaborative data is compressed, packaged, and encapsulated. Determine the business priority level based on the type of collaborative task; The modification probability value of the current collaborative data is obtained from the prediction model, and the prediction risk priority is calculated based on the confidence and lift of the prediction model related to the parameters. A comprehensive priority index is generated based on the business priority level and the predicted risk priority. The encapsulated collaborative data is then arranged in descending order of the comprehensive priority index to form an upload queue. Real-time monitoring of network quality status; dynamic adjustment of upload strategy based on network quality; use accelerated transmission protocol to upload collaborative data with high comprehensive priority index; cache collaborative data with medium and low comprehensive priority index to local queue and upload them in queue order. The uploaded data is recorded using a blockchain hash value, and the data consistency is verified through a consensus algorithm after the network is connected.

6. A task collaboration management system based on offline applications, characterized in that, include: The resource package acquisition module is used to determine the task collaboration type based on user information and task semantic parsing results, and acquire the corresponding collaborative task resource package. This includes: decompressing the collaborative task resource package to obtain a data model pre-set file, a business model pre-set file, and a collaborative form style pre-set file; automatically creating the data structure of the task collaboration system in the device's offline database based on the data model pre-set file, the data structure including master-slave table relationships and initialization parameters; automatically generating a business model object entity containing business rules in the device's offline application based on the business model pre-set file, the business rules including numerical validation logic and flow control rules; automatically creating a form interface for the collaborative task in the device's offline application based on the collaborative form style pre-set file, the form interface being associated with the data structure and the business model object entity; and loading the edge computing module to complete the generation of the task collaboration system environment. The environment initialization module is used to initialize and generate the task collaboration system environment based on the collaborative task resource package; A data early warning module is used to receive collaborative data uploaded by users through the task collaboration system environment, predict the modification probability of the collaborative data using a pre-trained prediction model in the collaborative task resource package, and issue early warnings and verifications for collaborative data with high-risk prediction results. The training method of the prediction model includes: extracting historical collaborative data and corresponding modification records from the collaborative task resource package or an offline database; the modification records include modification timestamps of historical collaborative data, modification frequencies of each parameter, and modification records of associated parameters; training a single-parameter modification frequency prediction model using an LSTM time series algorithm, wherein the LSTM algorithm weights the historical modification frequencies using a time decay factor; training the associated parameter modification records using an FP-Growth association rule algorithm to generate a parameter association modification prediction model; the FP-Growth algorithm mines the correlation between parameters using confidence and improvement metric indicators, and determines whether a strong correlation exists between parameters using preset trigger conditions; storing the single-parameter modification frequency prediction model and the parameter association modification prediction model on a local device for real-time prediction in offline mode. The data upload module is used to compress and package the collaborative data according to the collaborative task type, and to upload the packaged collaborative data through a priority scheduling mechanism when the network connection is restored.

7. A computer-readable storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, implements the method described in any one of claims 1-5.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the method of any one of claims 1-5.

Images