Task scheduling method and device

Abstract

The present invention provides a task scheduling method and device. The method includes: determining the mapping relationship between the task set and the task processing object set based on the problem description information of the task set and the skill domain information of the task processing object set; analyzing the mapping relationship between the task set and the task processing object set using an improved genetic model, and determining the global optimal task allocation strategy based on the fitness corresponding to the task processing object; wherein the fitness corresponds to the load factor of the task processing object and the residence time factor of the task assigned to the task processing object. The task scheduling method disclosed in the present invention can avoid falling into the local optimal task allocation strategy, balance the scheduling cost and scheduling efficiency, improve the stability and accuracy of the task scheduling allocation strategy, and improve the flow efficiency of customer problems, thereby optimizing the customer service experience.

Description

Technical Field

[0001] The present invention relates to the field of big data processing technology, and more particularly to a task scheduling method and apparatus, as well as an electronic device and a processor-readable storage medium. Background Art

[0002] In recent years, with the rapid development of the internet, an increasing number of applications have been enabled by network technology. Promptly addressing user feedback to improve the user experience has become a key research topic. In particular, with the rapid growth of enterprise service volumes, a wide variety of customer issues arise in daily service support. Enterprise service scenarios are often extremely complex. After-sales service scenarios face daily challenges posed by factors such as the sheer volume of services, diverse customer issues, and diverse technical experts (i.e., task handlers). Matching complex service scenarios with specialized technical experts and assigning numerous customer technical issues to designated service technical experts is an NP-hard problem. To efficiently resolve customer issues in complex service environments, scheduling systems face three key challenges: task processing efficiency, uneven technical expert load, and task response timeouts.

[0003] As the scale of problems continues to expand, traditional rule-based scheduling solutions, lacking a comprehensive approach, are clearly no longer sufficient. Furthermore, some technical experts face challenges with uneven task assignments and excessive workloads caused by continuous tasking. While genetic algorithm-based solutions can effectively balance the workload of technical experts, they suffer from slow solution speeds and unstable solutions. Therefore, designing a system that balances scheduling costs and efficiency, improves the flow of customer issues, and optimizes the customer service experience has become a key research topic for those skilled in the art. Summary of the Invention

[0004] To this end, the present invention provides a task scheduling method and device to solve the problem in the prior art of slow task allocation strategy determination and unstable process based on traditional genetic algorithm.

[0005] In a first aspect, the present invention provides a task scheduling method, comprising:

[0006] Determining a mapping relationship between the task set and the task processing object set based on the problem description information of the task set and the skill domain information of the task processing object set;

[0007] According to the mapping relationship between the task set and the task processing object set, an improved genetic model is used for analysis, and a global optimal task allocation strategy is determined based on the fitness corresponding to the task processing object; wherein the fitness corresponds to the load factor of the task processing object and the residence time factor of the task assigned to the task processing object.

[0008] In one embodiment, the task scheduling method further includes:

[0009] Before determining the mapping relationship, clustering the tasks in advance according to key description information in the tasks using a density-based clustering analysis model to generate corresponding task sets;

[0010] Wherein, each of the task sets corresponds to a category of problem description information;

[0011] Clustering the task processing objects using a K-means clustering analysis model based on the skill domain information of the task processing objects to generate a corresponding task processing object set;

[0012] Each task processing object in the task processing object set is used to match and process the task set corresponding to at least one category of the problem description information.

[0013] In one embodiment, according to the mapping relationship between the task set and the task processing object set, an improved genetic model is used for analysis, and a global optimal task allocation strategy is determined based on the fitness corresponding to the task processing objects, specifically including:

[0014] Based on the mapping relationship between the task set and the task processing object set, selecting a first task set and a second task set corresponding to a first task processing object and a second task processing object in the task processing object set, respectively; wherein the fitness corresponding to the first task processing object and the second task processing object are both lower than a preset fitness threshold;

[0015] After merging the first task set and the second task set, sorting the tasks from shortest to longest according to their processing time to generate a task list to be assigned; and determining the number information of each task in the task list based on the subscript of the task list;

[0016] Based on the numbering information and a preset greedy algorithm model, reallocate the odd-numbered tasks in the task list to the first task processing object and reallocate the even-numbered tasks in the task list to the second task processing object in accordance with a preset task allocation strategy; or, according to the task allocation strategy, allocate the even-numbered tasks in the task list to the first task processing object and allocate the odd-numbered tasks in the task list to the second task processing object;

[0017] If the average fitness value of the first task processing object and the average fitness value of the second task processing object are both lower than the average fitness value of the task processing object before reallocation, then continue the selection and reallocation iterative operation; if the average fitness value of the first task processing object and the average fitness value of the second task processing object are both higher than the average fitness value of the task processing object before reallocation, then retain the task allocation strategy;

[0018] When the reassignment iteration operation satisfies a preset stopping iteration rule, a global optimal task allocation strategy is determined based on the task allocation strategy.

[0019] In one embodiment, the task scheduling method further includes: determining the load factor based on the total working time of the task processing object in processing the total task and the standard working time of the task processing object.

[0020] In one embodiment, the task scheduling method further includes: determining a stay duration factor of the task based on a processing time of each task processed by the task processing object and a response time of each task processed by the task processing object.

[0021] In one embodiment, the fitness corresponds to the load factor of the task processing object and the residence time factor of the task assigned to the task processing object. Specifically, the smaller the load factor of the task processing object and the residence time factor of the task, the higher the fitness.

[0022] In a second aspect, the present invention further provides a task scheduling device, comprising:

[0023] A mapping relationship determining unit, configured to determine a mapping relationship between the task set and the task processing object set based on the problem description information of the task set and the skill domain information of the task processing object set;

[0024] A task allocation strategy determination unit is used to analyze the mapping relationship between the task set and the task processing object set using an improved genetic model, and determine the global optimal task allocation strategy based on the fitness corresponding to the task processing object; wherein the fitness corresponds to the load factor of the task processing object and the residence time factor of the task assigned to the task processing object.

[0025] In one embodiment, the task scheduling device further includes:

[0026] A task set generating unit is used to cluster the tasks using a density-based clustering analysis model according to key description information in the tasks before determining the mapping relationship, so as to generate a corresponding task set;

[0027] Wherein, each of the task sets corresponds to a category of problem description information;

[0028] A task processing object set generating unit, configured to cluster the task processing objects using a K-means clustering analysis model according to the skill domain information of the task processing objects, and generate a corresponding task processing object set;

[0029] Each task processing object in the task processing object set is used to match and process the task set corresponding to at least one category of the problem description information.

[0030] In one embodiment, the task allocation strategy determination unit is specifically configured to:

[0031] Based on the mapping relationship between the task set and the task processing object set, selecting a first task set and a second task set corresponding to a first task processing object and a second task processing object in the task processing object set, respectively; wherein the fitness corresponding to the first task processing object and the second task processing object are both lower than a preset fitness threshold;

[0032] After merging the first task set and the second task set, sorting the tasks from shortest to longest according to their processing time to generate a task list to be assigned; and determining the number information of each task in the task list based on the subscript of the task list;

[0033] Based on the numbering information and a preset greedy algorithm model, reallocate the odd-numbered tasks in the task list to the first task processing object and reallocate the even-numbered tasks in the task list to the second task processing object in accordance with a preset task allocation strategy; or, according to the task allocation strategy, allocate the even-numbered tasks in the task list to the first task processing object and allocate the odd-numbered tasks in the task list to the second task processing object;

[0034] If the average fitness value of the first task processing object and the average fitness value of the second task processing object are both lower than the average fitness value of the task processing object before reallocation, then continue the selection and reallocation iterative operation; if the average fitness value of the first task processing object and the average fitness value of the second task processing object are both higher than the average fitness value of the task processing object before reallocation, then retain the task allocation strategy;

[0035] When the reassignment iteration operation satisfies a preset stopping iteration rule, a global optimal task allocation strategy is determined based on the task allocation strategy.

[0036] In one embodiment, the task scheduling device further includes: a load factor determination unit for determining the load factor based on the total working time of the task processing object processing the total task and the standard working time of the task processing object.

[0037] In one embodiment, the task scheduling device further includes: a stay duration factor determination unit, which is used to determine the stay duration factor of the task based on the processing time of the task processing object for processing each task and the response time of the task processing object for processing each task.

[0038] In one embodiment, the fitness corresponds to the load factor of the task processing object and the residence time factor of the task assigned to the task processing object. Specifically, the smaller the load factor of the task processing object and the residence time factor of the task, the higher the fitness.

[0039] In a third aspect, the present invention further provides an electronic device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of the task scheduling method as described in any one of the above items are implemented.

[0040] In a fourth aspect, the present invention further provides a processor-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the task scheduling method as described in any one of the above items are implemented.

[0041] The task scheduling method described in the present invention can avoid falling into the local optimal task allocation strategy, balance the scheduling cost and scheduling efficiency, improve the stability and accuracy of the task scheduling allocation strategy, and improve the flow efficiency of customer issues, thereby optimizing the customer service experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following is a brief introduction to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0043] Figure 1 A flowchart of a task scheduling method provided by an embodiment of the present invention;

[0044] Figure 2 A schematic diagram of the structure of a task scheduling device provided by an embodiment of the present invention;

[0045] Figure 3A schematic diagram of the physical structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0046] To make the purpose, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0047] The present invention proposes a task scheduling method. First, a clustering algorithm is used to cluster tasks and task processing objects, and a mapping relationship between a task set and a task processing object set is established. Secondly, a genetic algorithm is used to solve the global optimal task allocation strategy. During the solution process, the genetic algorithm is improved to enhance the domain search capability of the genetic algorithm, thereby ensuring that the global optimal solution, that is, the global optimal task allocation strategy, can be better approximated, thereby improving the quality of understanding, and enhancing the problem handling efficiency and customer service experience.

[0048] The following describes in detail the embodiment of the task scheduling method according to the present invention. Figure 1 As shown in FIG, it is a flow chart of the task scheduling method provided by an embodiment of the present invention, and the specific implementation process includes the following steps:

[0049] Step 101: Determine a mapping relationship between the task set and the task processing object set based on the problem description information of the task set and the skill domain information of the task processing object set.

[0050] First of all, it should be noted that whether the technical field corresponding to the task processing object matches the problem classification of the task will directly affect the time it takes for the task processing object to process the task. The time taken by each task processing object to process the same task is different. There are three time nodes in task processing: task generation time -> start processing time -> processing end time. Among them, the period from task generation to the start of processing is called response time, and the period from the start of processing to the end of processing is called processing time. The total stay time of the task = response time + processing time = processing end time - task generation time. In order to ensure the service experience, each type of task has a maximum response time limit T, that is, the task needs to start processing within time T after it is generated, otherwise it will be counted as a service response timeout.

[0051] Before executing step 101, it is necessary to cluster the tasks in advance according to the key description information in the tasks using a density-based clustering analysis model (DBSCAN algorithm model) to generate a corresponding task set. According to the skill field information of the task processing object, the task processing objects are clustered using a K-means clustering analysis model (K-Means algorithm model) to generate a corresponding task processing object set. Each of the task sets corresponds to a category of problem description information. Each task processing object in the task processing object set is used to match and process the task set corresponding to at least one category of the problem description information. The task refers to a problem to be solved raised by a customer, and the task generated after the customer submits the problem to the system. The task processing object refers to a service technical expert, that is, a task processing personnel. The key description information refers to the key description information of the problem.

[0052] In addition, factors influencing the allocation target must be determined in advance. For example, the load factor is determined based on the total working time of the task processing object and the standard working time of the task processing object; and the task dwell time factor is determined based on the processing time of each task processed by the task processing object and the response time of each task processed by the task processing object.

[0053] During the specific implementation process, an allocation algorithm is designed to obtain a globally optimal task allocation strategy by assigning all unassigned tasks to task processing objects while satisfying the required constraints. The optimal task allocation strategy is one that minimizes the actual dwell time of all tasks (i.e., the dwell time factor cannot be too large) and does not overload individual task processing objects (i.e., the load factor cannot be too large) while satisfying the maximum response time constraint.

[0054] The load factor f of task processing object i i It can be calculated by the following formula:

[0055]

[0056]

[0057] Among them, T ij It represents the working time of task processing object i to process task j. Indicates the total working time of task processing object i in processing its own work tasks, t s Indicates the standard working time of the task processing object, i = 1, 2, ..., m, j = 1, 2, ..., n.

[0058] The duration factor s of task processing object i processing task j ij It can be obtained by the following formula:

[0059]

[0060] Where p ij The processing time for task processing object i to process task j, r ij The response time for task processing object i to process task j, where i = 1, 2, ..., m and j = 1, 2, ..., n.

[0061] Then, the optimal allocation problem to be solved can be described as:

[0062]

[0063] in, The maximum response time for task processing object i to process task j.

[0064] Furthermore, during the assignment process, tasks are clustered using the DBSCAN algorithm based on their key descriptive information to generate task sets. Task processing objects are clustered using the K-Means algorithm based on their skill domain information to generate task processing object sets. A mapping relationship is established between task sets and task processing object sets based on the key descriptive information of the task sets and the skill domain information of the task processing objects. Task sets mapped to the same task processing object set are merged to obtain a sub-problem set equal in number to the number of task processing object sets. Different clustering algorithms are used for tasks and task processing objects. Generally, the number of tasks is large, often tens or even hundreds of times the number of task processing objects. To ensure maximum task diversity during the clustering process, it is not advisable to limit the number of clusters. Therefore, the DBSCAN algorithm is used for task clustering. However, the number of task processing objects is generally small, and the areas of expertise of the task processing objects are relatively fixed. Therefore, the number of clusters can be limited based on actual conditions. Therefore, the K-Means algorithm is used to cluster the task processing objects.

[0065] Step 102: Analyze the mapping relationship between the task set and the task processing object set using an improved genetic model, and determine a global optimal task allocation strategy based on the fitness corresponding to the task processing objects.

[0066] In this step, first, based on the mapping relationship between the task set and the task processing object set, the first task set and the second task set corresponding to the first task processing object and the second task processing object in the task processing object set are selected. Wherein, the fitness corresponding to the first task processing object and the second task processing object are both lower than the preset fitness threshold; after merging the first task set and the second task set, they are sorted from short to long according to the task processing time to generate a task list to be assigned; and the numbering information of each task in the task list is determined based on the subscript of the task list. Then, based on the numbering information and the preset greedy algorithm model, according to the preset task allocation strategy, the odd-numbered tasks in the task list are reallocated to the first task processing object, and the even-numbered tasks in the task list are reallocated to the second task processing object; or, according to the task allocation strategy, the even-numbered tasks in the task list are allocated to the first task processing object, and the odd-numbered tasks in the task list are allocated to the second task processing object. If the average fitness value of the first task processing object and the average fitness value of the second task processing object are both lower than the average fitness value of the task processing object before reallocation, the selection and reallocation iteration operation is continued; if the average fitness value of the first task processing object and the average fitness value of the second task processing object are both higher than the average fitness value of the task processing object before reallocation, the task allocation strategy is retained. When the reallocation iteration operation satisfies the preset stopping iteration rule, the global optimal task allocation strategy is determined based on the task allocation strategy. The improved genetic model includes the greedy algorithm model.

[0067] The fitness corresponds to the load factor of the task processing object and the dwell time factor of the task assigned to the task processing object. That is, the smaller the load factor of the task processing object and the dwell time factor of the task, the higher the fitness.

[0068] In a specific embodiment, for each task set-task processing object set, an improved genetic model (i.e., an improved genetic algorithm) is used to solve the optimal allocation strategy for the subproblems, as follows:

[0069] (1) Initialize the population. Given the initial parameters of the genetic algorithm, the FirstFit algorithm can be used to solve the problem. Tasks are randomly assigned to task processing objects to obtain an initial solution. Each task processing object in the initial solution has an individual fitness, which is related to the load factor of the task processing object and the task residence time factor assigned to the task processing object. The smaller the sum of the load factor and the task residence time factor, the higher the fitness.

[0070] (2) Set the stopping rule of the genetic algorithm. The stopping rule includes two methods: the number of iterations and the solution accuracy. Determine whether the stopping rule is reached. If so, stop; otherwise, continue.

[0071] (3) Use the improved genetic algorithm to find the optimal solution to the subproblem:

[0072] (3.1) Select the operation to modify.

[0073] In the selection operation, in order to improve the overall fitness of the population, two individuals with lower fitness, namely the task processing objects, are selected first because the individuals with lower fitness have greater room and possibility for improvement, thereby reducing the blindness of the genetic algorithm search.

[0074] (3.2) Mutation operation modification.

[0075] In the specific implementation process, the mutation operation in the genetic algorithm is removed, thereby reducing the blindness of the genetic algorithm search.

[0076] (3.3) Cross-operation modification.

[0077] In the crossover operation, the traditional probability-oriented crossover operation is modified to further determine the search direction of the genetic algorithm. The goal of the crossover operation is to maximize the average fitness of the two selected individuals. The objective function to be solved is defined as:

[0078] ψ(n1,n2)=max(g(n1)+g(n2))

[0079] Among them, n1 and n2 are the two individuals selected by (3.3); g(n1) and g(n2) are the individual fitness of individuals n1 and n2.

[0080] In the specific implementation process, the tasks corresponding to the two selected individuals (i.e., the first task processing object and the second task processing object) are first extracted, merged, and sorted from shortest to longest by task processing time to generate a list of tasks to be assigned. Each task is numbered using the task list subscript. Further, using the preset greedy algorithm model, according to the preset task allocation strategy, odd-numbered tasks are assigned to individual 1 (i.e., the first task processing object) and even-numbered tasks are assigned to individual 2 (i.e., the second task processing object); or even-numbered tasks are assigned to individual 1 and odd-numbered tasks are assigned to individual 2. If the average individual fitness of both allocation strategies is lower than the average before reallocation, jump to (3.3) to continue the selection and reallocation iterative operation. Otherwise, retain the allocation strategy with the higher objective function.

[0081] (4) Return to (2) and determine whether the stopping condition is met. If so, stop the iteration; if not, continue the iteration.

[0082] Furthermore, in actual implementation, the improved genetic algorithm is used again to find the globally optimal task allocation strategy for the obtained solution to the task set-task processing object set subproblem, i.e., performing secondary optimization. The specific algorithm flow is consistent with the content of the above-mentioned specific embodiment and will not be repeated here.

[0083] In an embodiment of the present invention, the DBSCAN clustering algorithm is used to classify tasks, the K-Means algorithm is used to classify task processing objects, a mapping relationship is established from a task set to a task processing object set, solution constraints are defined, the scale of the solution space is reduced, and problem-solving efficiency is increased. The traditional genetic algorithm is improved, mainly including: modifying selection and crossover operations, removing mutation operations, clarifying the algorithm search direction, and improving search efficiency. A two-layer optimization algorithm is used, based on the good global search capability of the traditional genetic algorithm, to enhance the algorithm's neighborhood search capability, solving the problem that the genetic algorithm is prone to falling into local optimal solutions due to insufficient mutation.

[0084] The task scheduling method described in the embodiment of the present invention can avoid falling into the local optimal task allocation strategy, balance the scheduling cost and scheduling efficiency, improve the stability and accuracy of the task scheduling allocation strategy, and improve the flow efficiency of customer issues, thereby optimizing the customer service experience.

[0085] Corresponding to the task scheduling method provided above, the present invention also provides a task scheduling device located at the base station side. Since the embodiment of the device is similar to the above method embodiment, the description is relatively simple. For relevant details, please refer to the description of the above method embodiment. The embodiment of the task scheduling device described below is only illustrative. Please refer to Figure 2 As shown, it is a structural diagram of a task scheduling device provided by an embodiment of the present invention.

[0086] The task scheduling device described in the present invention specifically includes the following parts:

[0087] A mapping relationship determining unit 201 is configured to determine a mapping relationship between a task set and a task processing object set based on problem description information of the task set and skill domain information of the task processing object set;

[0088] The task allocation strategy determination unit 202 is configured to analyze the mapping relationship between the task set and the task processing object set using an improved genetic model and determine a global optimal task allocation strategy based on the fitness of the task processing objects. The fitness corresponds to the load factor of the task processing object and the residence time factor of the task assigned to the task processing object.

[0089] The task scheduling device described in the embodiment of the present invention can avoid falling into the local optimal task allocation strategy, balance the scheduling cost and scheduling efficiency, improve the stability and accuracy of the task scheduling allocation strategy, and improve the flow efficiency of customer issues, thereby optimizing the customer service experience.

[0090] Corresponding to the task scheduling method provided above, the present invention also provides an electronic device. Since the embodiment of the electronic device is similar to the embodiment of the method above, the description is relatively simple. For relevant details, please refer to the description of the embodiment of the method above. The electronic device described below is only for illustration. Figure 3 As shown, it is a schematic diagram of the physical structure of an electronic device disclosed in an embodiment of the present invention. The electronic device may include: a processor 301, a memory 302 and a communication bus 303, wherein the processor 301 and the memory 302 communicate with each other through the communication bus 303 and communicate with the outside through the communication interface 304. The processor 301 can call the logic instructions in the memory 302 to execute the task scheduling method. The method includes: determining the mapping relationship between the task set and the task processing object set based on the problem description information of the task set and the skill domain information of the task processing object set; based on the mapping relationship between the task set and the task processing object set, using an improved genetic model to analyze, and determining the global optimal task allocation strategy based on the fitness corresponding to the task processing object; wherein the fitness corresponds to the load factor of the task processing object and the stay time factor of the task assigned to the task processing object.

[0091] In addition, the logic instructions in the above-mentioned memory 302 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: a memory chip, a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program code.

[0092] On the other hand, an embodiment of the present invention further provides a computer program product, the computer program product comprising a computer program stored on a processor-readable storage medium, the computer program comprising program instructions, and when the program instructions are executed by a computer, the computer is able to execute the task scheduling method provided by each of the above method embodiments. The method comprises: determining a mapping relationship between the task set and the task processing object set based on the problem description information of the task set and the skill domain information of the task processing object set; analyzing the mapping relationship between the task set and the task processing object set using an improved genetic model, and determining a global optimal task allocation strategy based on the fitness corresponding to the task processing object; wherein the fitness corresponds to the load factor of the task processing object and the residence time factor of the task assigned to the task processing object.

[0093] In another aspect, an embodiment of the present invention further provides a processor-readable storage medium having a computer program stored thereon. When executed by a processor, the computer program is implemented to execute the task scheduling method provided in each of the above embodiments. The method includes: determining a mapping relationship between the task set and the task processing object set based on the problem description information of the task set and the skill domain information of the task processing object set; analyzing the mapping relationship between the task set and the task processing object set using an improved genetic model, and determining a global optimal task allocation strategy based on the fitness corresponding to the task processing object; wherein the fitness corresponds to the load factor of the task processing object and the residence time factor of the task assigned to the task processing object.

[0094] The processor-readable storage medium can be any available medium or data storage device that can be accessed by the processor, including but not limited to magnetic storage (such as floppy disks, hard disks, magnetic tapes, magneto-optical disks (MO)), optical storage (such as CDs, DVDs, BDs, HVDs, etc.), and semiconductor storage (such as ROMs, EPROMs, EEPROMs, non-volatile memories (NANDFLASH), solid-state drives (SSDs)), etc.

[0095] 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, i.e., they may be located in one location or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of the present embodiment. Persons of ordinary skill in the art will be able to understand and implement the present invention without inventive effort.

[0096] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, or of course, by hardware. Based on this understanding, the essence of the above technical solution or the part that contributes to the existing technology can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, an optical disk, etc., and includes a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or certain parts of the embodiments.

[0097] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A task scheduling method, characterized in that: include: Determining a mapping relationship between the task set and the task processing object set based on the problem description information of the task set and the skill domain information of the task processing object set; An improved genetic model is used to analyze the mapping relationship between the task set and the task processing object set, and a global optimal task allocation strategy is determined based on the fitness corresponding to the task processing object; wherein the fitness corresponds to the load factor of the task processing object and the residence time factor of the task assigned to the task processing object; According to the mapping relationship between the task set and the task processing object set, an improved genetic model is used for analysis, and a global optimal task allocation strategy is determined based on the fitness corresponding to the task processing objects, specifically including: Based on the mapping relationship between the task set and the task processing object set, selecting a first task set and a second task set corresponding to a first task processing object and a second task processing object in the task processing object set, respectively; wherein the fitness corresponding to the first task processing object and the second task processing object are both lower than a preset fitness threshold; After merging the first task set and the second task set, sorting the tasks from shortest to longest according to their processing time to generate a task list to be assigned; and determining the number information of each task in the task list based on the subscript of the task list; Based on the numbering information and a preset greedy algorithm model, reallocate the odd-numbered tasks in the task list to the first task processing object and reallocate the even-numbered tasks in the task list to the second task processing object in accordance with a preset task allocation strategy; or, according to the task allocation strategy, allocate the even-numbered tasks in the task list to the first task processing object and allocate the odd-numbered tasks in the task list to the second task processing object; If the average fitness value of the first task processing object and the average fitness value of the second task processing object are both lower than the average fitness value of the task processing object before reallocation, then continue the selection and reallocation iterative operation; if the average fitness value of the first task processing object and the average fitness value of the second task processing object are both higher than the average fitness value of the task processing object before reallocation, then retain the task allocation strategy; When the reassignment iteration operation satisfies a preset stopping iteration rule, a global optimal task allocation strategy is determined based on the task allocation strategy.

2. The task scheduling method according to claim 1, characterized in that: Also includes: Before determining the mapping relationship, clustering the tasks in advance according to key description information in the tasks using a density-based clustering analysis model to generate corresponding task sets; Wherein, each of the task sets corresponds to a category of problem description information; Clustering the task processing objects using a K-means clustering analysis model based on the skill domain information of the task processing objects to generate a corresponding task processing object set; Each task processing object in the task processing object set is used to match and process the task set corresponding to at least one category of the problem description information.

3. The task scheduling method according to claim 1, wherein: Also includes: The load factor is determined based on the total working time of the task processing object in processing the total tasks and the standard working time of the task processing object.

4. The task scheduling method according to claim 1, wherein: Also includes: A stay duration factor of the task is determined based on a processing duration of each task processed by the task processing object and a response duration of each task processed by the task processing object.

5. The task scheduling method according to claim 1, wherein: The fitness corresponds to the load factor of the task processing object and the residence time factor of the task assigned to the task processing object. Specifically, the smaller the load factor of the task processing object and the residence time factor of the task, the higher the fitness.

6. A task scheduling device, characterized in that: include: A mapping relationship determining unit, configured to determine a mapping relationship between the task set and the task processing object set based on the problem description information of the task set and the skill domain information of the task processing object set; a task allocation strategy determination unit, configured to analyze the mapping relationship between the task set and the task processing object set using an improved genetic model, and determine a global optimal task allocation strategy based on the fitness corresponding to the task processing object; wherein the fitness corresponds to the load factor of the task processing object and the residence time factor of the task assigned to the task processing object; According to the mapping relationship between the task set and the task processing object set, an improved genetic model is used for analysis, and a global optimal task allocation strategy is determined based on the fitness corresponding to the task processing objects, specifically including: Based on the mapping relationship between the task set and the task processing object set, selecting a first task set and a second task set corresponding to a first task processing object and a second task processing object in the task processing object set, respectively; wherein the fitness corresponding to the first task processing object and the second task processing object are both lower than a preset fitness threshold; After merging the first task set and the second task set, sorting the tasks from shortest to longest according to their processing time to generate a task list to be assigned; and determining the number information of each task in the task list based on the subscript of the task list; Based on the numbering information and a preset greedy algorithm model, reallocate the odd-numbered tasks in the task list to the first task processing object and reallocate the even-numbered tasks in the task list to the second task processing object in accordance with a preset task allocation strategy; or, according to the task allocation strategy, allocate the even-numbered tasks in the task list to the first task processing object and allocate the odd-numbered tasks in the task list to the second task processing object; If the average fitness value of the first task processing object and the average fitness value of the second task processing object are both lower than the average fitness value of the task processing object before reallocation, then continue the selection and reallocation iterative operation; if the average fitness value of the first task processing object and the average fitness value of the second task processing object are both higher than the average fitness value of the task processing object before reallocation, then retain the task allocation strategy; When the reassignment iteration operation satisfies a preset stopping iteration rule, a global optimal task allocation strategy is determined based on the task allocation strategy.

7. The task scheduling device according to claim 6, characterized in that: Also includes: A task set generating unit is used to cluster the tasks using a density-based clustering analysis model according to key description information in the tasks before determining the mapping relationship, so as to generate a corresponding task set; Wherein, each of the task sets corresponds to a category of problem description information; A task processing object set generating unit, configured to cluster the task processing objects using a K-means clustering analysis model according to the skill domain information of the task processing objects, and generate a corresponding task processing object set; Each task processing object in the task processing object set is used to match and process the task set corresponding to at least one category of the problem description information.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the steps of the task scheduling method according to any one of claims 1 to 5 are implemented.

9. A processor-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the task scheduling method according to any one of claims 1 to 5 are implemented.

Images