Party organization task multi-level distribution and management system

Abstract

The invention relates to the technical field of organization task management, in particular to a party organization task multi-level distribution and management system which comprises a task decomposition module used for analyzing tasks to obtain a task decomposition result which comprises a plurality of subtasks; the task hierarchy analysis module is used for performing hierarchy analysis on the sub-tasks of the task decomposition result and obtaining a plurality of hierarchical task groups according to the analysis result, and the hierarchical task groups comprise a plurality of sub-tasks of the same level; the task allocation module is used for allocating corresponding persons in charge to the sub-tasks of each hierarchical task group; and the task management module is used for updating and checking the execution condition of each task. The organization task multi-level distribution and management system provided by the invention can realize automatic reasonable distribution of multi-level organization tasks, and is suitable for distribution and management of task work of the multi-level organization.

Description

Technical Field

[0001] The present invention relates to the technical field of task management and scheduling, and specifically to a multi-level distribution and management system for party organization tasks. Background Art

[0002] In the daily work of various enterprises, organizations, and groups, task allocation and management is an important and complex task. The traditional task allocation method often relies on manual arrangements, which is not only inefficient, but also prone to problems such as unreasonable task allocation and difficulty in tracking progress. With the continuous expansion of organizational scale and the increasing complexity of work tasks, traditional management models are increasingly unable to meet the needs of efficient and accurate task management. Especially in a multi-level organizational structure, the superior organization needs to decompose the tasks and issue them to the grassroots organizations step by step. In this process, there is a lack of effective information technology support, resulting in low efficiency in task communication and execution, which affects the overall progress of the work.

[0003] Although some task management systems currently available on the market can achieve basic task allocation and tracking functions, these general systems are difficult to meet the special needs of organizations. For example, most systems only provide information processes, focusing on the electronic recording of data; they cannot achieve automatic and reasonable task allocation arrangements; although some systems can rely on historical allocation records to achieve a certain degree of intelligent automated task allocation, they have poor flexibility and cannot cope with different or unrecorded task types; in addition, the current system lacks optimization for the characteristics of multi-level organizations, and cannot effectively utilize the correlation and differences in the execution of tasks at different levels. There is a problem of low efficiency in task allocation. At the same time, the hierarchical targeting and correlation of task allocation are low, which leads to the problem of low allocation accuracy. Summary of the invention

[0004] The technical problem solved by the present invention is to provide a multi-level task allocation and management system for party organizations, which can solve the problems of low task allocation efficiency, low hierarchical targeting and correlation of task allocation existing in existing task allocation systems or methods, and is suitable for the allocation and management of task work in multi-level organizations.

[0005] The basic solution provided by the present invention is: A multi-level distribution and management system for party organization tasks, characterized by: A task decomposition module is used to analyze the task and obtain a task decomposition result, wherein the task decomposition result includes a plurality of subtasks; A task hierarchical analysis module is used to perform hierarchical analysis on the subtasks of the task decomposition result, and obtain a plurality of hierarchical task groups according to the analysis result, wherein the hierarchical task groups include a plurality of subtasks of the same level; Task assignment module, used to assign the subtasks of each hierarchical task group to the corresponding person in charge; The task management module is used to update and view the execution status of each task.

[0006] Further, the task allocation module includes: The allocation plan generation module is used to allocate the person in charge of the corresponding level according to the task level of the subtask, generate several allocation plans, and form an allocation plan set; The allocation scheme optimization module is used to optimize and iterate the allocation scheme set generated by the allocation scheme generation module by improving the multi-level ant lion optimization algorithm to generate the optimal allocation scheme; The allocation scheme optimization module includes: The scheme optimization and updating module is used to obtain the level allocation scheme set on each level from the allocation scheme set; according to the level from high to low, the level allocation scheme set of each level is processed by calling the level factor calculation module, the fitness calculation module, and the ant lion algorithm module; The hierarchical factor calculation module sets the upper-level task impact factor matrix and the upper-level personnel impact factor matrix to zero matrices when the current task level is the highest level; when the current task level is not the highest level, obtains the upper-level optimal allocation plan of the hierarchical allocation plan set of the upper level, and calculates the upper-level task impact factor matrix and the upper-level personnel impact factor matrix according to the upper-level optimal allocation plan:

[0007]

[0008] in, is the upper-level task impact factor matrix, It represents the nth task at the level level, dis() represents the distance between two tasks, and Normal() is the normalization operation;

[0009]

[0010] in, is the influence factor matrix of upper-level personnel, Represents the user corresponding to the nth subtask in the i-th allocation plan at the level level, represents the oth subtask of the upper layer's optimal allocation solution; The fitness calculation module is used to calculate the fitness of each allocation scheme at the current level and obtain the fitness score of each scheme:

[0011]

[0012]

[0013] in: is the fitness of the i-th allocation scheme at the level level; To obtain the maximum value, is the first weighted value, is the second weighted value; The total number of available personnel at the corresponding level of the organization. The total number of subtasks to be assigned at the level level; It is represented as the user corresponding to the jth subtask in the i-th allocation scheme at the level level. It is represented as the average time it takes for a user to process a single task corresponding to the jth subtask in the i-th allocation scheme at the level level; for The value of the kth row and jth column of the matrix, for The value of the kth row and jth column of the matrix; The total time for the mth person to process the i-th allocation plan; The time it takes for the mth person to handle the jth subtask in the i-th allocation plan; The antlion algorithm module is used to update the positions of antlions and ants according to the calculation results of the fitness calculation module. The updated positions of ants and antlions constitute a new set of level allocation schemes for the current level; the new level allocation schemes for all levels constitute a new allocation scheme set; The iteration module is used to determine whether the exit condition is met. If so, the scheme combination corresponding to the position of the elite ant lion in each level allocation scheme is selected as the optimal allocation scheme. Otherwise, the scheme optimization update module and the iteration module are repeatedly called to process the new allocation scheme set.

[0014] Furthermore, the allocation scheme optimization module also includes: The scheme screening module is used to determine whether each allocation scheme in the allocation scheme set meets the boundary conditions before the scheme optimization and updating module processes the allocation scheme set, and remove the allocation schemes that do not meet the boundary conditions from the allocation scheme set. The boundary conditions are:

[0015] in, The maximum processing time of the task.

[0016] Furthermore, the Antlion algorithm module includes: The ant lion position update module is used to update the positions of the ant lion and the ant according to the calculation results of the fitness calculation module, and update the position of the ant lion to the position of the ant with higher fitness, wherein the allocation scheme with the highest fitness is used as the position of the elite ant lion; The ant position update module adopts a roulette strategy to select target ant lions for ants; the ants are allowed to randomly walk according to the positions of target ant lions and elite ant lions, and the positions of the ants are updated.

[0017] Furthermore, the ant position update module first updates the minimum value of the ant g-th dimension variable at the t-th iteration according to the position of the antlion. and the maximum value of the g-th dimension variable at the t-th iteration :

[0018]

[0019] in, is the minimum value of all variables in the tth iteration; is the maximum value of all variables in the tth iteration; is the position of the jth antlion at the tth iteration; I is the ratio, whose value is , T is the maximum number of iterations, ω is adjusted according to the current iteration, specifically, ω=2 when t>0.1 T, ω=3 when t>0.5 T, ω=4 when t>0.75 T, ω=5 when t>0.9 T, ω=6 when t>0.95 T; Then the position of the g-th ant in the t-th iteration is , randomly wandering around antlions and elite antlions:

[0020] in, is the minimum value of the random walk of the g-th dimension variable, is the maximum value of the random walk of the g-th dimension variable; Then update the ant's position:

[0021] in, is the position of the random walk around the antlion selected according to the roulette strategy at the tth iteration, is the position of the random walk around the elite antlion in the tth iteration.

[0022] Furthermore, the task decomposition module includes: Task input module, used for users to input task orders; The semantic parsing and extraction module is used to parse the input task book and extract the text summary; A prompt word construction module is used to match the corresponding prompt word template according to the text summary, and construct the prompt word content according to the text summary and the prompt word template; The generative AI module is used to output task decomposition results based on the input prompt word content.

[0023] Furthermore, the semantic parsing and extraction module includes: Extractive text summarization model, used to extract text from task books; The extractive text summarization model includes an input layer, a word classification layer, a sentence classification layer and an output layer according to the data flow direction; the sentence classification layer corresponds to multiple word classification layers, a bidirectional RNN connection structure is used between the sentence classification layers, a bidirectional RNN connection structure is used between multiple word classification layers corresponding to the same sentence classification layer, and the output layer uses a sigmoid activation function for classification output; The type matching module is used to classify the extracted text to obtain task-related information, which includes task type, task goal, task description, and task time.

[0024] Furthermore, the prompt word construction module includes: A template matching module, used to match a prompt word template from a prompt word template library according to a task type; The prompt word construction module is used to fill task-related information into the matching prompt word template to form the prompt word content.

[0025] Furthermore, the task level analysis module includes: A task level prediction module is used to predict the level involved in the subtasks through a task level prediction model according to the task information of the subtasks in the task decomposition results; The task level determination module takes the highest level involved in the subtask as the level of the subtask; The task grouping module is used to divide the subtasks into multiple hierarchical task groups according to the levels of each subtask.

[0026] Furthermore, the task management module includes: The task progress acquisition module is used to obtain the actual execution progress of the subtasks corresponding to each task; The task progress warning module obtains the progress plan of the subtask according to the task information of the subtask, obtains the progress difference according to the progress plan and the actual execution progress obtained by the task progress acquisition module, and determines whether the progress difference is greater than a threshold. If it is greater than the threshold, a task progress warning message is generated.

[0027] The beneficial effects of the present invention are: 1. The text summary extraction algorithm combined with the generative AI model can realize task decomposition and allocation in scenarios without historical records and human participation, which improves the flexibility and versatility of the system. Compared with the existing manual classification or historical record allocation method, task decomposition can be more scientific.

[0028] 2. This paper innovatively proposes the ML-ALO improved multi-level antlion optimization algorithm. For the multi-level task allocation problem of this application, it fully considers the possible dependencies or mutual influences between different levels, and optimizes the lower-level task allocation through the upper-level task influence factor matrix and the upper-level personnel influence factor matrix, thereby achieving the overall optimization effect; improving the hierarchical targeting and hierarchical correlation of task allocation, and improving the accuracy of task allocation.

[0029] 3. Through the influence of the upper layer on the lower layer task allocation, the optional space of the lower layer tasks is reduced, the operation efficiency and convergence speed of the entire algorithm are improved, and the efficiency of task allocation is improved. In addition, the multi-level ant lion optimization algorithm can realize multi-level parallel operation, further improving the operation efficiency of the optimization algorithm. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 This is a logical block diagram of a multi-level distribution and management system for party organization tasks in an embodiment of the present invention. DETAILED DESCRIPTION

[0031] The following is further described in detail through specific implementation methods: Embodiment 1 A multi-level distribution and management system for party organization tasks in this embodiment is used for task allocation, task management and progress control within the organization. The system is based on WEB and native development as a whole, and adopts a B / S and C / S hybrid architecture. It mainly includes four parts: server backend system, data center, user end and management end. The backend service system is used to provide various business logic services for the operation and management of the entire system. The data center provides data storage services and third-party data access services for the entire system. The user end and management end are used for users and managers to operate system functions.

[0032] The backend service system uses a containerization solution based on the Kubernetes engine, combined with a microservice framework built with SpringBoot. Each business module in the backend server system is implemented using independent microservices. Microservices can call each other through agreed interfaces and protocols, and multiple microservices can be aggregated into a unified microservice. Since each microservice is deployed independently, it can be deployed iteratively quickly, and the appropriate virtual machine or container can be selected according to the service requirements, and the hardware that best matches the service resource requirements can be used. The overall application is decomposed into manageable modules and services, and individual services can be developed faster, understood and maintained more easily.

[0033] In this embodiment, the business functions of the background service system mainly include: organization management subsystem, user management subsystem, task allocation and management subsystem, visualization large screen subsystem, report export subsystem and push service subsystem.

[0034] Among them, the organization management subsystem is mainly used to set up and manage the system architecture, department settings, and responsible persons of the overall organization; the user management subsystem is mainly used to add, delete, and edit information of specific personnel within the organization; the task allocation and management subsystem is used to realize multi-level allocation of tasks and progress control; the visualization large-screen subsystem is used to count and visualize task data, and the report export subsystem is used to automatically generate task or project reports; the push service subsystem is used to push system information to the user end and the management end, such as task allocation information, progress alarm information, organizational structure adjustment information, etc.

[0035] like Figure 1 As shown, the task allocation and management subsystem includes: The task decomposition module is used to analyze the task and obtain the task decomposition result, which includes several subtasks. In this embodiment, the task decomposition module receives the task information input by the management end. The task information is the overall task, including the description of the entire task, time limit, target, budget amount, technical route, location, personnel scope, etc. In other embodiments of the present application, it may not be limited to these fields. The task information usually covers the various contents required for the project establishment report of various activities in the organization. In this embodiment, the data structure corresponding to the task information is stored in a JSON string. The background can configure the various attribute dimensions of the task information and display it through the front-end page, which is convenient for realizing customized field control, improving the flexibility and versatility of the system, and adapting to different organizations or enterprises. The task decomposition module pushes the received task information to the corresponding user end. The user end can be a person in the organization who is specifically responsible for project or task arrangement, or a person designated by the management end or a person included in the task information. These can be pre-configured in the system. After the task decomposition module receives the task information, it pushes it to the corresponding user end according to the pre-configuration or task information, and then obtains the task decomposition result input by the user end for saving.

[0036] The task level analysis module is used to perform level analysis on the subtasks of the task decomposition result, and obtain multiple hierarchical task groups according to the analysis results, wherein the hierarchical task groups include several subtasks of the same level; due to the complexity of different tasks, the hierarchical levels of the personnel involved are different. After the task is decomposed into multiple subtasks, it is necessary to analyze the level of the subtasks according to the content of the subtasks. In this embodiment, for each subtask, the project level analysis and division are performed according to the budget and number of personnel involved in the subtask. For example, the personnel of an organization are divided into 5 levels according to the job level, and the subtasks are divided into five levels according to the budget amount: less than 100,000 yuan, 100,000-300,000 yuan, 300,000-500,000 yuan, 500,000-1 million yuan, and more than 1 million yuan. The number of project personnel is divided into five levels according to less than 5 people, 5-20 people, 20-50 people, 50-80 people, and more than 80 people. For each subtask, the corresponding level is matched according to the budget amount and the number of personnel, and the higher level of the two is adopted.

[0037] The task assignment module is used to assign the subtasks of each hierarchical task group to the corresponding person in charge. The task assignment module specifically includes: The allocation scheme generation module is used to allocate the person in charge of the corresponding level according to the task level of the subtask, generate several allocation schemes, and form an allocation scheme set. In this embodiment, the generation of the initial allocation scheme is mainly random, that is, the person in charge of the corresponding level is directly and randomly allocated according to the task level of the subtask, until all tasks of all levels are allocated, that is, an allocation scheme is formed; this is repeated many times to obtain several initial allocation schemes, which constitute an allocation scheme set. In this implementation, X=[ ] represents a combination of allocation schemes, where Represents a set of level allocation schemes for the level level. , Represents the i-th allocation plan for tasks at level level.

[0038] The allocation scheme optimization module is used to optimize and iterate the allocation scheme set generated by the allocation scheme generation module by improving the multi-level ant lion optimization algorithm to generate the optimal allocation scheme; the allocation scheme optimization module includes a scheme screening module, a scheme optimization and updating module and an iteration module.

[0039] The scheme screening module is used to determine whether each allocation scheme in the allocation scheme set meets the boundary conditions before the scheme optimization and updating module processes the allocation scheme set, and remove the allocation schemes that do not meet the boundary conditions from the allocation scheme set. The boundary conditions are:

[0040]

[0041]

[0042] in, is the maximum processing time of the task, The total number of available personnel at the corresponding level of the organization. The total number of subtasks to be assigned at the level level; It is represented as the user corresponding to the jth subtask in the i-th allocation scheme at the level level. It is represented as the average time it takes for a user to process a single task corresponding to the jth subtask in the i-th allocation scheme at the level level; for The value of the kth row and jth column of the matrix, for The value of the kth row and jth column of the matrix; The total time for the mth person to process the i-th allocation plan; The time it takes for the mth person to handle the jth subtask in the i-th allocation plan.

[0043] The scheme optimization update module is used to obtain the level allocation scheme set on each level from the allocation scheme set X. ; According to the level from high to low, the level allocation scheme set for each level Call the hierarchical factor calculation module, fitness calculation module, and antlion algorithm module for processing; The hierarchical factor calculation module sets the upper-level task impact factor matrix and the upper-level personnel impact factor matrix to zero matrices when the current task level is the highest level; when the current task level is not the highest level, obtains the upper-level optimal allocation plan of the hierarchical allocation plan set of the upper level, and calculates the upper-level task impact factor matrix and the upper-level personnel impact factor matrix according to the upper-level optimal allocation plan:

[0044]

[0045] in, is the upper-level task impact factor matrix, It represents the nth task under the level level, dis() is the distance calculation, and Normal() is the normalization operation. In this embodiment, dis() uses Euclidean distance to perform distance calculation. In other embodiments of the present application, cosine similarity, Manhattan distance, etc. can also be used to calculate the distance between two tasks.

[0046]

[0047]

[0048] in, is the influence factor matrix of upper-level personnel, Represents the user corresponding to the nth subtask in the i-th allocation plan at the level level, represents the oth subtask of the upper layer's optimal allocation solution; The fitness calculation module is used to calculate the fitness of each allocation scheme at the current level and obtain the fitness score of each scheme:

[0049] in: is the fitness of the i-th allocation scheme at the level level; To obtain the maximum value, is the first weighted value, is the second weighted value; The antlion algorithm module is used to update the positions of antlions and ants according to the calculation results of the fitness calculation module, and the updated positions of the ants and the positions of the antlions constitute a new level allocation scheme set for the current level; the new level allocation scheme sets for all levels constitute a new allocation scheme set; The Antlion algorithm module includes: The ant lion position update module is used to update the positions of the ant lion and the ant according to the calculation results of the fitness calculation module, and update the position of the ant lion to the position of the ant with higher fitness, wherein the allocation scheme with the highest fitness is used as the position of the elite ant lion; The ant position update module adopts a roulette strategy to select target ant lions for ants; the ants are allowed to randomly walk according to the positions of target ant lions and elite ant lions, and the positions of the ants are updated.

[0050] Specifically, first, according to the position of the antlion, the minimum value of its g-th dimension variable at the t-th iteration is updated according to the following formula: and the maximum value of the g-th dimension variable at the t-th iteration .

[0051]

[0052]

[0053] in, is the minimum value of all variables in the tth iteration; is the maximum value of all variables in the tth iteration; is the position of the jth antlion at the tth iteration; I is the ratio, whose value is , T is the maximum number of iterations, ω is a constant defined according to the current iteration, ω=2 when t>0.1 T, ω=3 when t>0.5 T, ω=4 when t>0.75 T, ω=5 when t>0.9 T, and ω=6 when t>0.95 T.

[0054] Then the position of the g-th ant in the t-th iteration is , randomly walk around the ant lion and elite ant lion according to the following formula.

[0055]

[0056] in, is the minimum value of the random walk of the g-th dimension variable, is the maximum value of the random walk of the g-th dimension variable.

[0057] Finally, take the average value according to the following formula to update the ant's position.

[0058]

[0059] in, is the position of the random walk around the antlion selected according to the roulette strategy at the tth iteration, is the position of the random walk around the elite antlion in the tth iteration.

[0060] The iteration module is used to determine whether the exit condition is met. If so, the solution combination corresponding to the position of the elite ant lion in each level allocation plan is selected as the optimal allocation plan. If not, the solution screening module, solution optimization and update module, and iteration module are repeatedly called to process the new allocation plan set. Then in this implementation, the exit condition is that the number of iterations reaches a preset value, and the specific preset number of iterations is 500 times.

[0061] In this implementation, the overall iterative loop is performed after all levels have been optimized once. In other embodiments of the present application, each level can be executed in a parallel iterative manner. That is, when each level has the data of its upper level, the upper-level task influence factor matrix and the upper-level personnel influence factor matrix can be calculated. Then, the next iteration of its upper-level task allocation can run in parallel with the current iteration of the lower-level task allocation.

[0062] The task management module is used to update and view the execution status of each task. The task management module includes: The task progress acquisition module is used to obtain the actual execution progress of the subtasks corresponding to each task; The task progress warning module obtains the progress plan of the subtask according to the task information of the subtask, obtains the progress difference according to the progress plan and the actual execution progress obtained by the task progress acquisition module, and determines whether the progress difference is greater than a threshold. If it is greater than the threshold, a task progress warning message is generated.

[0063] Embodiment 2 The difference between this embodiment and the first embodiment is that in this embodiment, the task decomposition module includes: The task input module is used for users to input task orders. In this implementation, it provides upload entrances for text, voice, image and files. Users can input task content through various multimedia methods such as text description, voice description, file upload, etc. The task input module integrates text recognition, image recognition, document content extraction, voice recognition and other modules to comprehensively obtain user input.

[0064] The semantic parsing and extraction module is used to parse the input task book and extract the text summary. Specifically, the semantic parsing and extraction module includes: The extractive text summary model is used to extract text from the task book; in this embodiment, the extractive text summary model adopts a text summary extractor based on a bidirectional RNN network, which includes an input layer, a word classification layer, a sentence classification layer and an output layer according to the data flow; the sentence classification layer corresponds to multiple word classification layers, and a bidirectional RNN connection structure is used between the sentence classification layers, and a bidirectional RNN connection structure is used between multiple word classification layers corresponding to the same sentence classification layer, and the output layer uses a sigmoid activation function for classification output. In other embodiments of the present application, other neural network structures can also be used to realize the extraction and processing of text summaries.

[0065] The type matching module is used to classify the extracted text and obtain task-related information, including task type, task goal, task description, budget amount, task time, staff size, task plan, etc. Since the extracted text is scattered, it is necessary to identify and fill in the content according to the category to realize the identification of task information.

[0066] A prompt word construction module is used to match the corresponding prompt word template according to the text summary, and construct the prompt word content according to the text summary and the prompt word template; the prompt word construction module includes: A template matching module, used to match a prompt word template from a prompt word template library according to a task type; The template management module is used for users to maintain and manage prompt word templates. In this implementation, various prompt word templates are preset in the system, and users can manually add, edit and delete prompt word templates. Each template has corresponding matching keywords and applicable task types.

[0067] The prompt word construction module is used to fill task-related information into the matching prompt word template to form the prompt word content.

[0068] The generative AI module is used to output the task decomposition results according to the input prompt word content. In this implementation, the generative AI module constructs a complete training and test set by inputting and outputting each manually decomposed task, and obtains a generative AI model dedicated to task decomposition and completes the deployment through training based on the ChatGLM-6B open source model. The generative AI module mainly obtains prompt words and transmits the prompt words to the generative AI model, and returns the output data of the model. In other embodiments of the present application, other existing AI models or AI interfaces of third-party platforms can also be used.

[0069] Embodiment 3 The difference between this embodiment and the first embodiment is that, in this embodiment, the task level analysis module includes a task level prediction module, which is used to predict the level involved in the subtask through the task level prediction model according to the task information of the subtask in the task decomposition result; Specifically, the task-level prediction model is implemented using a CNN convolutional neural network, which includes an input layer, a convolutional layer, a pooling layer, and a fully connected layer. The input layer converts the information description of the task into a two-dimensional vector matrix. In this implementation, the information of all tasks is arranged in a fixed format length to form a two-dimensional vector. The convolutional layer includes multiple convolution kernels with different radii for feature extraction of two-dimensional vector data. The pooling layer is used to pool and reduce the dimensionality of the data. In this implementation, the average pooling method is adopted, and then the task-level results are classified and output through a softmax fully connected layer. By obtaining historical task-level data, a training set for training is formed. After training the above-mentioned convolutional neural network, a task-level prediction model is obtained to achieve task-level prediction.

[0070] The task level determination module takes the highest level involved in the subtask as the level of the subtask; The task grouping module is used to divide the subtasks into multiple hierarchical task groups according to the levels of each subtask.

[0071] The above are only embodiments of the present invention. Common knowledge such as the known specific structures and characteristics in the scheme is not described in detail here. Ordinary technicians in the relevant field know all the common technical knowledge in the technical field to which the invention belongs before the application date or priority date, can obtain all the existing technologies in the field, and have the ability to apply conventional experimental means before that date. Ordinary technicians in the relevant field can improve and implement this scheme in combination with their own abilities under the enlightenment given by this application. Some typical known structures or known methods should not become obstacles for ordinary technicians in the relevant field to implement this application. It should be pointed out that for those skilled in the art, without departing from the structure of the present invention, several deformations and improvements can be made, which should also be regarded as the scope of protection of the present invention, and these will not affect the effect of the implementation of the present invention and the practicality of the patent. The scope of protection required by this application shall be based on the content of its claims, and the specific implementation methods and other records in the specification can be used to interpret the content of the claims.

Claims

1. A multi-level distribution and management system for party organization tasks, characterized by: include: A task decomposition module is used to analyze the task and obtain a task decomposition result, wherein the task decomposition result includes a plurality of subtasks; A task hierarchical analysis module is used to perform hierarchical analysis on the subtasks of the task decomposition result, and obtain a plurality of hierarchical task groups according to the analysis result, wherein the hierarchical task groups include a plurality of subtasks of the same level; The task assignment module is used to assign the subtasks of each hierarchical task group to the corresponding person in charge; Task management module, used to update and view the execution status of each task; The task allocation module includes: The allocation plan generation module is used to allocate the person in charge of the corresponding level according to the task level of the subtask, generate several allocation plans, and form an allocation plan set; An allocation plan optimization module is used to optimize and iterate the allocation plan set generated by the allocation plan generation module through an optimization algorithm to generate an optimal allocation plan; The optimization algorithm is an improved multi-level ant lion optimization algorithm, and the allocation scheme optimization module includes: The scheme optimization and updating module is used to obtain the level allocation scheme set on each level from the allocation scheme set; according to the level from high to low, the level allocation scheme set of each level is processed by calling the level factor calculation module, the fitness calculation module, and the ant lion algorithm module; The hierarchical factor calculation module sets the upper-level task impact factor matrix and the upper-level personnel impact factor matrix to zero matrices when the current task level is the highest level; when the current task level is not the highest level, obtains the upper-level optimal allocation plan of the hierarchical allocation plan set of the upper level, and calculates the upper-level task impact factor matrix and the upper-level personnel impact factor matrix according to the upper-level optimal allocation plan: in, is the upper-level task impact factor matrix, It represents the nth task at the level level, dis() represents the distance between two tasks, and Normal() is the normalization operation; in, is the influence factor matrix of upper-level personnel, Represents the user corresponding to the nth subtask in the i-th allocation plan at the level level, represents the oth subtask of the upper layer's optimal allocation solution; The fitness calculation module is used to calculate the fitness of each allocation scheme at the current level and obtain the fitness score of each scheme: in: is the fitness of the i-th allocation scheme at the level level; To obtain the maximum value, is the first weighted value, is the second weighted value; The total number of available personnel at the corresponding level of the organization. The total number of subtasks to be assigned at the level level; It is represented as the user corresponding to the jth subtask in the i-th allocation scheme at the level level. It is represented as the average time it takes for a user to process a single task corresponding to the jth subtask in the i-th allocation scheme at the level level. for The value of the kth row and jth column of the matrix, for The value of the kth row and jth column of the matrix; The total time for the mth person to process the i-th allocation plan; The time it takes for the mth person to handle the jth subtask in the i-th allocation plan; The antlion algorithm module is used to update the positions of antlions and ants according to the calculation results of the fitness calculation module. The updated positions of ants and antlions constitute a new set of level allocation schemes for the current level; the new level allocation schemes for all levels constitute a new allocation scheme set; The iteration module is used to determine whether the exit condition is met. If so, the combination of plans corresponding to the position of the elite ant lion in each level allocation plan is selected as the optimal allocation plan. Otherwise, the plan optimization update module and the iteration module are repeatedly called to process the new allocation plan set. When each level obtains the upper optimal allocation plan of the level allocation plan set of the previous level and calculates the upper task influence factor matrix and the upper personnel influence factor matrix, the next iteration of its upper task allocation and the current iteration of the current level task allocation run in parallel.

2. A multi-level distribution and management system for party organization tasks according to claim 1, characterized in that: The allocation scheme optimization module also includes: The scheme screening module is used to determine whether each allocation scheme in the allocation scheme set meets the boundary conditions before the scheme optimization and updating module processes the allocation scheme set, and remove the allocation schemes that do not meet the boundary conditions from the allocation scheme set. The boundary conditions are: in, is the maximum processing time of the task, To take the minimum value.

3. A multi-level distribution and management system for party organization tasks according to claim 2, characterized in that: The Antlion algorithm module includes: The ant lion position update module is used to update the positions of the ant lion and the ant according to the calculation results of the fitness calculation module, and update the position of the ant lion to the position of the ant with higher fitness, wherein the allocation scheme with the highest fitness is used as the position of the elite ant lion; The ant position update module adopts a roulette strategy to select target ant lions for ants; the ants are allowed to randomly walk according to the positions of target ant lions and elite ant lions, and the positions of the ants are updated.

4. A multi-level distribution and management system for party organization tasks according to claim 1, characterized in that: The task decomposition module includes: Task input module, used for users to input task orders; The semantic parsing and extraction module is used to parse the input task book and extract the text summary; A prompt word construction module is used to match the corresponding prompt word template according to the text summary, and construct the prompt word content according to the text summary and the prompt word template; The generative AI module is used to output task decomposition results based on the input prompt word content.

5. A multi-level distribution and management system for party organization tasks according to claim 4, characterized in that: The semantic parsing and extraction module comprises: Extractive text summarization model, used to extract text from task books; The extractive text summarization model includes an input layer, a word classification layer, a sentence classification layer and an output layer according to the data flow direction; the sentence classification layer corresponds to multiple word classification layers, a bidirectional RNN connection structure is used between the sentence classification layers, a bidirectional RNN connection structure is used between multiple word classification layers corresponding to the same sentence classification layer, and the output layer uses a sigmoid activation function for classification output; The type matching module is used to classify the extracted text to obtain task-related information, which includes task type, task goal, task description, and task time.

6. A multi-level distribution and management system for party organization tasks according to claim 5, characterized in that: The prompt word building module includes: A template matching module, used to match a prompt word template from a prompt word template library according to a task type; The prompt word construction module is used to fill task-related information into the matching prompt word template to form the prompt word content.

7. A multi-level distribution and management system for party organization tasks according to claim 1, characterized in that: Task level analysis module, including: A task level prediction module is used to predict the level involved in the subtasks through a task level prediction model according to the task information of the subtasks in the task decomposition results; The task level determination module takes the highest level involved in the subtask as the level of the subtask; The task grouping module is used to divide the subtasks into multiple hierarchical task groups according to the levels of each subtask.

8. A multi-level distribution and management system for party organization tasks according to claim 1, characterized in that: The task management module includes: The task progress acquisition module is used to obtain the actual execution progress of the subtasks corresponding to each task; The task progress warning module obtains the progress plan of the subtask according to the task information of the subtask, obtains the progress difference according to the progress plan and the actual execution progress obtained by the task progress acquisition module, and determines whether the progress difference is greater than a threshold. If it is greater than the threshold, a task progress warning message is generated.

Images