Atomic Capability Recommendation Method, Device, and Storage Medium for Process Orchestration

Abstract

The present application discloses an atomic capability recommendation method, device, and storage medium for process orchestration. The method includes: obtaining a first job flow to be orchestrated by a first user, and determining a first user attribute, first historical scoring behavior data, and a first job flow attribute; comparing the number of first atomic capabilities scored by the first user with a preset quantity threshold; when the number of first atomic capabilities is less than the preset quantity threshold, recommending a plurality of target atomic capabilities for orchestrating the first job flow to the first user according to the first user attribute and the first job flow attribute; otherwise, recommending a plurality of target atomic capabilities for orchestrating the first job flow to the first user according to the first user attribute, the first job flow attribute, and the first scores of the first user for each atomic capability. The present application solves the technical problem that when using a process orchestration system to orchestrate a job flow in the related art, it is difficult to select the most suitable atomic capabilities for each node of the job flow from a vast number of atomic capabilities.

Description

Technical Field

[0001] The present application relates to the technical field of process orchestration, and in particular, to a method, device, and storage medium for recommending atomic capabilities for process orchestration. Background Art

[0002] As the process orchestration development mode is becoming more and more popular among enterprises, enterprises are more and more inclined to use process orchestration systems for business development and operation and maintenance development. Whether it is a process orchestration system deployed privately in an enterprise or a public cloud process orchestration system, the number of its users is relatively large. With the construction of the process orchestration system, the number of atomic capabilities will also increase. When users use the process orchestration system to build a workflow, it is difficult to select the most needed atomic capabilities for the current node from a large number of atomic capabilities.

[0003] Therefore, related technical personnel generally recommend atomic capabilities in the following ways: searching for atomic capabilities by name; selecting fewer atomic capabilities by tags or classifications. However, the above-mentioned recommendation methods still have the following drawbacks: they cannot recommend atomic capabilities based on the historical behavior of the current user; they cannot recommend atomic capabilities based on the user attributes of the current user; they cannot recommend atomic capabilities based on the workflow attributes of the current workflow; they cannot recommend atomic capabilities based on the content of the current workflow.

[0004] For the above problems, no effective solution has been proposed yet. Summary of the Invention

[0005] Embodiments of the present application provide a method, device, and storage medium for recommending atomic capabilities for process orchestration, so as to at least solve the technical problem that it is difficult to select the most suitable atomic capabilities for each node of the workflow from a large number of atomic capabilities when using a process orchestration system to orchestrate a workflow in related technologies.

[0006] According to one aspect of the embodiments of the present application, a method for recommending atomic capabilities for process orchestration is provided, including: obtaining a first workflow to be orchestrated by a first user, and determining a first user attribute, first historical scoring behavior data of the first user, and a first workflow attribute of the first workflow, where the first historical scoring behavior data includes: a first score of the first user for each atomic capability in the atomic capability library, and the number of first atomic capabilities scored by the first user; comparing the number of first atomic capabilities scored by the first user with a preset quantity threshold; when the number of first atomic capabilities is less than the preset quantity threshold, recommending a plurality of target atomic capabilities for orchestrating the first workflow to the first user according to the first user attribute and the first workflow attribute; when the number of atomic capabilities is not less than the preset quantity threshold, recommending a plurality of target atomic capabilities for orchestrating the first workflow to the first user according to the first user attribute, the first workflow attribute, and the first score of the first user for each atomic capability.

[0007] Optionally, obtain the first job flow to be orchestrated by the first user, and determine the first user attribute, the first historical scoring behavior data of the first user, and the first job flow attribute of the first job flow, including: obtaining the first job flow to be orchestrated by the first user in the process orchestration system, and determining the first user attribute of the first user and the first job flow attribute of the first job flow, where the first user attribute includes at least one of the following: user role, user department, user position, and the first job flow attribute includes at least one of the following: process purpose, process grouping, process label; for each atomic ability in the atomic ability library, obtain the scoring behavior of the first user for the atomic ability, and determine the first score of the first user for the atomic ability according to the scoring behavior and the preset behavior weight, where the scoring behavior includes at least one of the following: the first user favorites the atomic ability, any job flow orchestrated by the first user uses the atomic ability, the first user views the atomic ability; summarize the first scores of the first user for each atomic ability, and determine the number of the first atomic abilities scored by the first user.

[0008] Optionally, the method for determining the preset quantity threshold includes: determining the number of atomic abilities scored by multiple users in the process orchestration system, and determining the median and average of the array composed of the number of atomic abilities scored by each user; calculating the average of the median and the average, and using the average as the preset quantity threshold.

[0009] Optionally, the method further includes: creating a first type of recommendation module, where the first type of recommendation module is used to analyze the historical scoring behavior data of each user in the process orchestration system to obtain the first recommendation coefficient matrix of each user for each atomic ability in the atomic ability library; creating a second type of recommendation module, where the second type of recommendation module is used to analyze the current job flow attribute of the current job flow to be orchestrated by the current user in the process orchestration system, the current user attribute, and the second user attributes of multiple other second users and the second job flow attributes of multiple other second job flows to obtain the atomic ability similarity matrix of each atomic ability in the atomic ability library; creating a third type of recommendation module, where the third type of recommendation module is used to analyze the first atomic ability of the first i working nodes in the current job flow to obtain the second recommendation coefficient matrix of each atomic ability in the atomic ability library, where i is a positive integer greater than 1.

[0010] Optionally, the analysis process of the first type of recommendation module includes: determining, by the first type of recommendation module, the second scores of each user for each atomic capability in the atomic capability library; constructing a score matrix based on the second scores of each user for each atomic capability in the atomic capability library, calculating the first similarity between any two atomic capabilities based on the second scores of each user for each atomic capability in the atomic capability library, and constructing an atomic capability similarity matrix from the first similarity between any two atomic capabilities; multiplying the score matrix by the atomic capability similarity matrix to obtain the first recommendation coefficient matrix of each user for each atomic capability in the atomic capability library.

[0011] Optionally, the analysis process of the second type of recommendation module includes: calculating, by the second type of recommendation module, the second similarity between the current user attribute of the current user and the second user attributes of each second user, and constructing a user similarity matrix based on the second similarity; converting the user similarity matrix into a job flow similarity weighted coefficient matrix for characterizing the similarity of the creating users of the job flow; calculating the third similarity between the current job flow attribute and each second job flow attribute, constructing a job flow similarity matrix based on the third similarity, and determining a job flow weighted similarity matrix based on the job flow similarity matrix and the job flow similarity weighted coefficient matrix; determining the job flow weighted similarity corresponding to each of the multiple second job flows where each atomic capability in the atomic capability library is located based on the job flow weighted similarity matrix, and determining the fourth similarity of the atomic capability based on the average value of the job flow weighted similarities corresponding to the multiple second job flows; constructing an atomic capability similarity matrix based on the fourth similarity of each atomic capability.

[0012] Optionally, calculating the second similarity between the current user attribute of the current user and the second user attribute of each second user by the second type of recommendation module includes: determining, by the second type of recommendation module, the number of identical user attributes that the current user and each second user have, and the total number of user attributes that the current user and each second user have; taking the quotient of the number of identical user attributes and the total number of user attributes as the second similarity between the current user attribute and the second user attribute of each second user.

[0013] Optionally, calculating the third similarity between the current job flow attribute and each second job flow attribute includes: determining, by the second type of recommendation module, the number of identical job flow attributes that the current job flow and each second job flow have, and the total number of job flow attributes that the current job flow and each second job flow have; taking the quotient of the number of identical job flow attributes and the total number of job flow attributes as the third similarity between the current job flow attribute and each second job flow attribute.

[0014] Optionally, the analysis process of the third type of recommendation module includes: extracting, through the third type of recommendation module, the first upstream sub-process of the i-th working node that has configured the first atomic capability in the current job flow, and the second upstream sub-processes of each working node that has configured the second atomic capability in multiple other second job flows; respectively constructing multiple first adjacency matrices of the first upstream sub-process and the multiple second upstream sub-processes, and multiple second adjacency matrices of each second upstream sub-process and the first upstream sub-process; for each second upstream sub-process, calculating the fifth similarity between the first upstream sub-process and the second upstream sub-process based on the first adjacency matrix of the first upstream sub-process and the second upstream sub-process and the second adjacency matrix of the second upstream sub-process and the first upstream sub-process, and determining the first recommendation coefficient of the second upstream sub-process according to the fifth similarity; for any second atomic capability in the second upstream sub-process, determining the second recommendation coefficient of the second atomic capability according to the first recommendation coefficients of the multiple second upstream sub-processes where the second atomic capability is located; and forming a second recommendation coefficient matrix from the second recommendation coefficients corresponding to each second atomic capability.

[0015] Optionally, respectively constructing multiple first adjacency matrices of the first upstream sub-process and the multiple second upstream sub-processes, and multiple second adjacency matrices of each second upstream sub-process and the first upstream sub-process, includes: for each second upstream sub-process, dividing the first upstream sub-process and the second upstream sub-process into the first type of atomic capability, the second type of atomic capability, and the third type of atomic capability through the third type of recommendation module, where the first type of atomic capability is used to reflect the unique atomic capability of the first upstream sub-process, the second type of atomic capability is used to reflect the common atomic capability of the first upstream sub-process and the second upstream sub-process, and the third type of atomic capability is used to reflect the unique atomic capability of the second upstream sub-process; using the first type of atomic capability, the second type of atomic capability, and the third type of atomic capability as rows and columns respectively, and respectively constructing the first adjacency matrix of the first upstream sub-process and the second upstream sub-process, and the second adjacency matrix of the second upstream sub-process and the first upstream sub-process.

[0016] Optionally, calculating a fifth similarity between the first upstream subprocess and the second upstream subprocess based on a first adjacency matrix of the first upstream subprocess and the second upstream subprocess and a second adjacency matrix of the second upstream subprocess and the first upstream subprocess includes: calculating a first Hadamard product of the first adjacency matrix of the first upstream subprocess and the second upstream subprocess and the second adjacency matrix of the second upstream subprocess and the first upstream subprocess by a third type of recommendation module and multiplying the result by 2 to obtain a similarity numerator matrix, and taking the square root of the sum of each first matrix element in the similarity numerator matrix to obtain a similarity numerator; calculating a second Hadamard product of the second adjacency matrix itself, combining it with the first Hadamard product to obtain a similarity denominator matrix, and taking the square root of the sum of each second matrix element in the similarity denominator matrix to obtain a similarity denominator; calculating the fifth similarity between the first upstream subprocess and the second upstream subprocess based on the similarity numerator and the similarity denominator.

[0017] Optionally, calculating a first recommendation coefficient of the second upstream subprocess according to the fifth similarity includes: determining the execution success rate of the second upstream subprocess by a third type of recommendation module; determining the first recommendation coefficient of the second upstream subprocess based on the product of the fifth similarity and the execution success rate.

[0018] Optionally, determining a second recommendation coefficient of the second atomic capability according to the first recommendation coefficients of multiple second upstream subprocesses where the second atomic capability is located includes: taking the result obtained by taking the square root of the sum of squares of the first recommendation coefficients of multiple second upstream subprocesses where the second atomic capability is located by a third type of recommendation module as the second recommendation coefficient of the second atomic capability.

[0019] Optionally, recommending multiple target atomic capabilities for orchestrating the first job flow to a first user according to the first user attribute and the first job flow attribute includes: invoking a second type of recommendation module to analyze the first user attribute and the first job flow attribute to obtain a first atomic capability similarity matrix, and recommending a target atomic capability of the first job node of the first job flow to the first user according to the first atomic capability similarity matrix; invoking the second type of recommendation module and the third type of recommendation module to analyze the first user attribute, the first job flow attribute, and the target atomic capabilities of the first i job nodes to obtain a first atomic capability similarity matrix and a second target recommendation coefficient matrix, and recommending a target atomic capability corresponding to the (i + 1)-th job node of the first job flow to the first user according to the first atomic capability similarity matrix and the second target recommendation coefficient matrix.

[0020] Optionally, recommending to the first user the target atomic capability corresponding to the (i + 1)-th job node of the first job flow according to the first atomic capability similarity matrix and the second target recommendation coefficient matrix includes: recommending to the first user the target atomic capability corresponding to the (i + 1)-th job node of the first job flow according to the first atomic capability similarity matrix, the weight factor of the first atomic capability similarity matrix, and the second target recommendation coefficient matrix.

[0021] Optionally, recommending to the first user multiple target atomic capabilities for orchestrating the first job flow according to the first user attribute, the first job flow attribute, and the first user's ratings of each atomic capability includes: invoking a first type of recommendation module and a second type of recommendation module to analyze the first user attribute, the first job flow attribute, and the first ratings of each atomic capability by the first user to obtain a first target recommendation coefficient matrix and a first atomic capability similarity matrix, and recommending to the first user the target atomic capability of the first job node of the first job flow according to the first target recommendation coefficient matrix and the first atomic capability similarity matrix; invoking the first type of recommendation module, the second type of recommendation module, and a third type of recommendation module to analyze the first user attribute, the first job flow attribute, the first ratings of each atomic capability by the first user, and the target atomic capabilities of the previous i job nodes to obtain a first target recommendation coefficient matrix, a first atomic capability similarity matrix, and a second target recommendation coefficient matrix, and recommending to the first user the target atomic capability corresponding to the (i + 1)-th job node of the first job flow according to the first target recommendation coefficient matrix, the first atomic capability similarity matrix, and the second target recommendation coefficient matrix.

[0022] Optionally, recommending to the first user the target atomic capability of the first job node of the first job flow according to the first target recommendation coefficient matrix and the first atomic capability similarity matrix includes: recommending to the first user the target atomic capability of the first job node of the first job flow according to the first atomic capability similarity matrix, the weight factor of the first atomic capability similarity matrix, and the first target recommendation coefficient matrix.

[0023] Optionally, recommending to the first user the target atomic capability corresponding to the (i + 1)-th job node of the first job flow according to the first target recommendation coefficient matrix, the first atomic capability similarity matrix, and the second target recommendation coefficient matrix includes: recommending to the first user the target atomic capability corresponding to the (i + 1)-th job node of the first job flow according to the first atomic capability similarity matrix, the weight factor of the first atomic capability similarity matrix, the first target recommendation coefficient matrix, and the second target recommendation coefficient matrix.

[0024] According to another aspect of the embodiments of the present application, there is also provided an atomic capability recommendation device for process orchestration, including: a determination module, configured to obtain a first job flow to be orchestrated by a first user, and determine a first user attribute, first historical scoring behavior data of the first user, and a first job flow attribute of the first job flow, where the first historical scoring behavior data includes: a first score of the first user for each atomic capability in the atomic capability library, and the number of first atomic capabilities scored by the first user; a comparison module, configured to compare the number of first atomic capabilities scored by the first user with a preset quantity threshold; a first recommendation module, configured to recommend a plurality of target atomic capabilities for orchestrating the first job flow to the first user according to the first user attribute and the first job flow attribute when the number of first atomic capabilities is less than the preset quantity threshold; a second recommendation module, configured to recommend a plurality of target atomic capabilities for orchestrating the first job flow to the first user according to the first user attribute, the first job flow attribute, and the first score of the first user for each atomic capability when the number of atomic capabilities is not less than the preset quantity threshold.

[0025] According to another aspect of the embodiments of the present application, there is also provided a non-volatile storage medium, which includes a stored computer program, where the device where the non-volatile storage medium is located executes the above-mentioned atomic capability recommendation method for process orchestration by running the computer program.

[0026] In the embodiments of the present application, obtain a first job flow to be orchestrated by a first user, and determine a first user attribute, first historical scoring behavior data of the first user, and a first job flow attribute of the first job flow, where the first historical scoring behavior data includes: a first score of the first user for each atomic capability in the atomic capability library, and the number of first atomic capabilities scored by the first user; compare the number of first atomic capabilities scored by the first user with a preset quantity threshold; when the number of first atomic capabilities is less than the preset quantity threshold, recommend a plurality of target atomic capabilities for orchestrating the first job flow to the first user according to the first user attribute and the first job flow attribute; when the number of atomic capabilities is not less than the preset quantity threshold, recommend a plurality of target atomic capabilities for orchestrating the first job flow to the first user according to the first user attribute, the first job flow attribute, and the first score of the first user for each atomic capability.

[0027] In the above technical solution, according to whether the number of first atomic capabilities scored by the first user reaches a preset quantity threshold, the atomic capability recommendation is divided into two cases: when the number of first atomic capabilities is insufficient, the process orchestration system can recommend multiple target atomic capabilities for orchestrating the first job flow to the first user according to the first user attributes and the first job flow attributes; when the number of first atomic capabilities is sufficient, the process orchestration system can recommend multiple target atomic capabilities for orchestrating the first job flow to the first user according to the first user attributes, the first job flow attributes, and the first scores of the first user for each atomic capability. The process orchestration system makes full use of various types of information to achieve the accuracy, quality, and robustness of atomic capability recommendation, thereby solving the technical problem that it is difficult to select the most suitable atomic capabilities for each node of the job flow from a large number of atomic capabilities when using the process orchestration system to orchestrate the job flow in the related art. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation of the present application. In the drawings:

[0029] Figure 1 is a hardware structure block diagram of a computer terminal for implementing an atomic capability recommendation method for process orchestration according to an embodiment of the present application;

[0030] Figure 2 is an optional process schematic diagram for implementing an atomic capability recommendation method for process orchestration according to an embodiment of the present application;

[0031] Figure 3 is an optional process schematic diagram for implementing step S206 according to an embodiment of the present application;

[0032] Figure 4 is an optional process schematic diagram for implementing step S208 according to an embodiment of the present application;

[0033] Figure 5 is a schematic structural diagram of an optional device for implementing an atomic capability recommendation method for process orchestration according to an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0034] In order to enable those skilled in the art to better understand the solution of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without making creative efforts shall fall within the protection scope of the present application.

[0035] It should be noted that the terms "first", "second", etc. in the description, claims and the above-mentioned drawings of this application are used to distinguish similar objects, and do not necessarily describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances, so that the embodiments of this application described here can be implemented in an order other than those illustrated or described here. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device comprising a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0036] In addition, the relevant information (including but not limited to user equipment information, user personal information, etc.) and data (including but not limited to data for display, data for analysis, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. For example, an interface is set between this system and relevant users or institutions. Before obtaining relevant information, a request for acquisition needs to be sent to the aforementioned user or institution through the interface, and after receiving the consent information feedback from the aforementioned user or institution, the relevant information can be obtained.

[0037] To better understand the embodiments of this application, the technical terms involved in the embodiments of this application are explained as follows:

[0038] Atomic capability is a standardized abstraction of network resource capabilities. It is the smallest unit with certain service attributes and is used to orchestrate network services. The following contents can all be defined as atomic capabilities: abstracting the service functions of network devices and the IT capabilities of network management to support the indivisible and common configuration units or elements of network servers; a configuration switch with a single function, which can be turned on or off on network devices, and after being turned on, a specific network function can be provided, such as video compression; a license with a single function, and after loading this license, a network server with a specific function can be provided. Atomic capabilities are located in the "network atomic capabilities" layer of the hierarchical framework. Atomic capabilities have the following characteristics: (1) Atomic capabilities have service attributes and are used to support the development of network services; (2) Atomic capabilities have atomicity, which describes a specific network function and is indivisible; (3) There may be dependencies between atomic capabilities; (4) Atomic capabilities can be reused and can be orchestrated to form different network services.

[0039] A job flow refers to the realization of several operations to complete a certain task by orchestrating atomic capabilities. The execution sequence between atomic capabilities is represented by connection lines.

[0040] A working node is a specific step in a job flow, and each node can select only one atomic capability.

[0041] The Hadamard Product, also translated as the Hadamard multiplication, is a binary operation. Its inputs are two matrices of the same shape, and the output is a matrix of the same shape, where each element at each position is equal to the product of the elements at the same position in the two input matrices. That is, if two matrices A and B have the same dimension m*n, then their Hadamard product A⊙B is a matrix of the same dimension, and its element value is: (A⊙B) ij =(A) ij (B) ij . It should be noted that for matrices with different dimensions (an m*n matrix and a p*q matrix, and m≠p, n≠q), the Hadamard product is not defined.

[0042] Embodiment 1

[0043] Currently, when users build a work process using a process orchestration system, they usually filter atomic capabilities based on names, tags, or classifications, which makes it difficult for users to select the atomic capabilities most needed for the current node from a large number of atomic capabilities.

[0044] To solve this problem, the embodiments of the present application provide an embodiment of a method for recommending atomic capabilities for process orchestration, aiming to recommend atomic capabilities for specific users, specific job flows, and specific nodes, effectively reducing the time for filtering atomic capabilities, improving the utilization rate of high-quality atomic capabilities, and thus enhancing development efficiency and code quality.

[0045] It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. And although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than here.

[0046] The method embodiments provided by the embodiments of the present application can be executed on a mobile terminal, a computer terminal, or a similar computing device. Figure 1 Shows a hardware structure block diagram of a computer terminal for implementing a method for recommending atomic capabilities for process orchestration. As Figure 1As shown, the computer terminal 10 (or mobile device 10) may include one or more processors 102 (illustrated as 102a, 102b, ……, 102n in the figure) (the processor 102 may include, but is not limited to, processing devices such as a microprocessor MCU or a field programmable gate array FPGA), a memory 104 for storing data, and a transmission device 106 for communication functions. In addition, it may further include: a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of the BUS bus), a network interface, a power supply, and / or a camera. Those of ordinary skill in the art can understand that Figure 1 the structure shown is only schematic and does not limit the structure of the above-mentioned electronic device. For example, the computer terminal 10 may further include more or fewer components than Figure 1 shown in, or have a different configuration from Figure 1 that shown.

[0047] It should be noted that the above one or more processors 102 and / or other data processing circuits may generally be referred to as "data processing circuits" herein. The data processing circuit may be embodied in software, hardware, firmware, or any combination thereof, in whole or in part. In addition, the data processing circuit may be a single independent processing module, or be incorporated in whole or in part into any one of the other elements in the computer terminal 10. As involved in the embodiments of the present application, the data processing circuit is a kind of processor control (such as the selection of a variable resistance terminal path connected to an interface).

[0048] The memory 104 may be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the atomic capability recommendation method for process choreography in the embodiments of the present application. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, that is, implements the atomic capability recommendation method for process choreography of the above-mentioned application program. The memory 104 may include a high-speed random access memory, and may further include a non-volatile memory, such as one or more magnetic storage devices, a flash memory, or other non-volatile solid-state memories. In some instances, the memory 104 may further include a memory remotely disposed relative to the processor 102, and these remote memories may be connected to the computer terminal 10 through a network. Examples of the above network include, but are not limited to, the Internet, an enterprise intranet, a local area network, a mobile communication network, and combinations thereof.

[0049] The transmission device 106 is used to receive or send data via a network. Specific examples of the above-mentioned network may include a wireless network provided by the communication provider of the computer terminal 10. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 can be a Radio Frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0050] The display can be, for example, a touch-screen liquid crystal display (LCD), which enables users to interact with the user interface of the computer terminal 10.

[0051] In the above operating environment, Figure 2 is a schematic flowchart of an optional atomic capability recommendation method for process orchestration according to an embodiment of the present application, as Figure 2 shown, the method at least includes steps S202 - S208, where:

[0052] Step S202, obtain the first job flow to be orchestrated by the first user, and determine the first user attribute, the first historical scoring behavior data of the first user, and the first job flow attribute of the first job flow.

[0053] In the technical solution provided in step S202, the process orchestration system obtains the first job flow to be orchestrated by the first user, and determines the first user attribute of the first user and the first historical scoring behavior data of the first user. Among them, the above-mentioned first historical scoring behavior data includes: the first score of the first user for each atomic capability in the atomic capability library, and the number of the first atomic capabilities scored by the first user. In addition, the process orchestration system also needs to obtain the first job flow attribute of the first job flow. Subsequently, the process orchestration system can recommend the atomic capabilities required for each work node of the first job flow based on this information.

[0054] Step S204, compare the number of the first atomic capabilities scored by the first user with a preset quantity threshold.

[0055] The purpose of this step is to determine whether the number of the first atomic capabilities scored by the first user in history is sufficient. Among them, if the number of the first atomic capabilities is not sufficient, the process orchestration system uses the first job flow attribute and the first user attribute to recommend atomic capabilities; if the number of the first atomic capabilities is sufficient, the process orchestration system combines the first score of the first user for the atomic capabilities to recommend atomic capabilities.

[0056] Step S206, when the number of first atomic capabilities is less than the preset quantity threshold, recommend multiple target atomic capabilities for orchestrating the first job flow to the first user according to the first user attribute and the first job flow attribute.

[0057] Step S208, when the number of atomic capabilities is not less than the preset quantity threshold, recommend multiple target atomic capabilities for orchestrating the first job flow to the first user according to the first user attribute, the first job flow attribute, and the first user's first score for each atomic capability.

[0058] It should be noted that the above steps S206 and S208 do not distinguish the order of execution during the process.

[0059] Based on the solution defined in the above steps S202 to S208, it can be known that in the embodiment, obtain the first job flow to be orchestrated by the first user, and determine the first user attribute of the first user, the first historical scoring behavior data, and the first job flow attribute of the first job flow. Among them, the first historical scoring behavior data includes: the first user's first score for each atomic capability in the atomic capability library, and the number of first atomic capabilities that the first user has scored; compare the number of first atomic capabilities that the first user has scored with the preset quantity threshold; when the number of first atomic capabilities is less than the preset quantity threshold, recommend multiple target atomic capabilities for orchestrating the first job flow to the first user according to the first user attribute and the first job flow attribute; when the number of atomic capabilities is not less than the preset quantity threshold, recommend multiple target atomic capabilities for orchestrating the first job flow to the first user according to the first user attribute, the first job flow attribute, and the first user's first score for each atomic capability.

[0060] Thus, through the technical solution of the embodiment of the present application, the process orchestration system makes full use of various types of information to achieve the accuracy, quality, and robustness of atomic capability recommendation, and further solves the technical problem that it is difficult to select the most suitable atomic capabilities for each node of the job flow from a large number of atomic capabilities when using the process orchestration system to orchestrate the job flow in the related art.

[0061] The above method of this embodiment will be further introduced below.

[0062] As an optional implementation manner, in the technical solution provided in the above step S202, the method may include the following steps S2021 - S2023, including:

[0063] Step S2021, obtain the first job flow to be orchestrated by the first user in the process orchestration system, and determine the first user attribute of the first user and the first job flow attribute of the first job flow.

[0064] Among them, the first user attribute includes at least one of the following: user role, user department to which the user belongs, user position, and the first job flow attribute includes at least one of the following: process purpose, process grouping, process label.

[0065] Step S2022: For each atomic ability in the atomic ability library, obtain the scoring behavior of the first user for the atomic ability, and determine the first score of the first user for the atomic ability according to the scoring behavior and the preset behavior weight.

[0066] In the technical solution provided in the above step S2022, the scoring behavior of the first user for the atomic ability includes at least one of the following: the first user favorites the atomic ability, any job flow arranged by the first user uses the atomic ability, and the first user views the atomic ability. That is to say, if the user favorites the atomic ability, the score of the user for the atomic ability increases; if the user uses the atomic ability when arranging a certain job flow, the score will increase by one each time it is used; if the user views the relevant content of the atomic ability, the score of the user for the atomic ability increases. Therefore, the expression for the score of user i for atomic ability a can be written as:

[0067] r a,i = α1 + α2x + α3y

[0068] Among them, α1 represents the weight of the favorite behavior, α2 represents the weight of the use behavior, x represents the number of times user i uses atomic ability a, α3 represents the weight of the view behavior, and y represents the number of times user i views atomic ability a.

[0069] Step S2023: Aggregate the first scores of the first user for each atomic ability, and determine the number of the first atomic abilities scored by the first user.

[0070] Furthermore, in order to determine whether the historical scoring behavior data of the first user is sufficient, the process orchestration system can also compare the number of the first atomic abilities scored by the first user determined in the above step S2023 with a preset quantity threshold. The purpose of doing this is that when the historical scoring behavior data of the first user is insufficient, if the first score of the first user for the atomic ability is used for atomic ability recommendation, the recommendation coefficients of each atomic ability cannot be accurately calculated. Therefore, in the embodiments of the present application, it is necessary to compare the size relationship between the number of the first atomic abilities and the preset quantity threshold.

[0071] Specifically, the preset quantity threshold can be determined in the following manner, including: First, determine the number of atomic abilities scored by multiple users in the process orchestration system, and determine the median and average of the array composed of the number of atomic abilities scored by multiple users; then, calculate the average of the median and the average, and use the average as the preset quantity threshold. Among them, the above multiple users refer to users with process orchestration behavior.

[0072] According to the size relationship between the number of first atomic capabilities rated by the first user and a preset quantity threshold, the atomic capability recommendation by the process orchestration system to the first user can be divided into the above two recommendation processes of step S206 and step S208. To facilitate the process orchestration system to quickly recommend atomic capabilities, embodiments of the present application also create multiple recommendation modules based on multiple types of recommendation information respectively, so that the process orchestration system can use these recommendation modules alone or in combination to accurately recommend atomic capabilities.

[0073] As an optional implementation manner, a first type of recommendation module is created. Among them, the first type of recommendation module is used to analyze the historical rating behavior data of each user in the process orchestration system to obtain a first recommendation coefficient matrix of each atomic capability in the atomic capability library for each user; a second type of recommendation module is created. Among them, the second type of recommendation module is used to analyze the current job flow attributes of the current job flow to be orchestrated by the current user in the process orchestration system, the current user attributes, as well as the second user attributes of multiple other second users and the second job flow attributes of multiple other second job flows to obtain an atomic capability similarity matrix of each atomic capability in the atomic capability library; a third type of recommendation module is created. Among them, the third type of recommendation module is used to analyze the first i configured first atomic capabilities in the current job flow to obtain a second recommendation coefficient matrix of each atomic capability in the atomic capability library, where i is a positive integer greater than 1.

[0074] Among them, the above-mentioned each module can be understood as follows: the above-mentioned first type of recommendation module recommends atomic capabilities based on the historical behavior of the current user; the second type of recommendation module recommends atomic capabilities based on the current job flow and the current user attributes; the third type of recommendation module recommends atomic capabilities based on the configured work nodes of the current job flow. And the above-mentioned second job flow can be understood as the job flow that has been orchestrated in the process orchestration system.

[0075] Specifically, regarding the above-mentioned first type of recommendation module, the algorithm used by this module is item-based collaborative filtering, where the item is an atomic capability. Considering that the user's rating of the atomic capability is based on objective facts, there is no problem of user subjective tendency. At the same time, in order to adapt to the sparsity problem of less rating data, the first type of recommendation module uses cosine similarity to calculate the similarity between each atomic capability. The following will illustrate the specific analysis process of the first type of recommendation module through the following steps S11 - S13, including:

[0076] Step S11, determine the second rating of each atomic capability in the atomic capability library for each user through the first type of recommendation module;

[0077] Step S12: Construct a scoring matrix based on the second scores of each atomic ability in the atomic ability library by the user, and calculate the first similarity between any two atomic abilities according to the second scores of each atomic ability in the atomic ability library by the user, and construct an atomic ability similarity matrix from the first similarities between any two atomic abilities;

[0078] Step S13: Multiply the scoring matrix by the atomic ability similarity matrix to obtain the first recommendation coefficient matrix of the user for each atomic ability in the atomic ability library.

[0079] In the above embodiment, first, the second scores of each atomic ability in the atomic ability library by each user are determined by the first type of recommendation module. Among them, the process of the user scoring each atomic ability has been described in detail in the above step S2022, so it will not be elaborated here.

[0080] Then, use this second score to generate a scoring matrix of the current user for all atomic abilities, which can be denoted as:

[0081] R i ={R a,1} m×1

[0082] where i represents the user, and R a,1 represents the second score of user i for atomic ability a, and m represents the number of atomic abilities.

[0083] At the same time, according to the second scores of each atomic ability in the atomic ability library by the user, use cosine similarity to calculate the first similarity between every two atomic abilities. The expression of this first similarity is:

[0084]

[0085] where a and b respectively represent two atomic abilities, r a,i and r b,i respectively represent the second scores of user i for atomic ability a and atomic ability b, and n represents the number of users in the process choreography system. Use the first similarity s(a, b) between every two atomic abilities to construct an atomic ability similarity matrix, denoted as S={S a,b} m×m where S a,b is the first similarity between atomic ability a and atomic ability b. For example, the atomic ability similarity matrix S can be expressed in the form of Table 1 below:

[0086] Table 1

[0087] b＝1 b＝2 … b = m a＝1 <![CDATA[S 1,1 > <![CDATA[S 1,2 > … <![CDATA[S 1,m > a＝2 <![CDATA[S 2,1 > <![CDATA[S 2,2 > … <![CDATA[S 2,m > … … … … … a = m <![CDATA[S m,1 > <![CDATA[S m,2 > … <![CDATA[S m,m >

[0088] Finally, multiply the atomic ability similarity matrix S and the scoring matrix R iMultiply to obtain the first recommendation coefficient matrix P of all atomic capabilities for user i i , and its expression is:

[0089] P i = S × R i = {S a,b} m×m × {R a,1} m×1 = {P a,1} m×1

[0090] where P a,1 represents the recommendation coefficient of user i for atomic capability a.

[0091] Regarding the second type of recommendation module, since various attributes such as labels, purposes, and groups can be set when creating a job flow, and users also have various attributes such as roles, positions, and departments, the second type of recommendation module can recommend atomic capabilities according to this information through the following steps S21 - S25, including:

[0092] Step S21, calculate the second similarity between the current user attributes of the current user and the second user attributes of each second user through the second type of recommendation module, and construct a user similarity matrix based on the second similarity.

[0093] Optionally, in the technical solution provided in step S21 above, the second type of recommendation module can also determine the second similarity through the following method:

[0094] Step S211, determine the number of identical user attributes that the current user and each second user have through the second type of recommendation module, as well as the total number of user attributes that the current user and each second user have;

[0095] Step S212, use the quotient of the number of identical user attributes and the total number of user attributes as the second similarity between the current user attributes and the second user attributes of each second user.

[0096] In the above embodiment, the user attributes of the current user i are denoted as B0, and the second user attributes of each other second user are denoted as B j , where j represents the user serial number of other second users. Therefore, the number of identical user attributes that the current user i and each second user have can be expressed as: |B0 ∩ B j |, and the total number of user attributes that the current user i and each second user have is expressed as: |B0 ∪ B j |, and then calculate the second similarity between the current user attributes and the second user attributes of each second user through the following formula:

[0097]

[0098] Finally, by repeating the above steps S211 - S212, the user similarity matrix between the current user i and each of the other second users can be obtained, denoted as {SB j,1} m×1 .

[0099] Step S22: Convert the user similarity matrix into a job flow similarity weighted coefficient matrix for characterizing the similarity of the creating users of the job flows.

[0100] The above step S22 can be understood as follows: Since a job flow is created by a certain user, the user similarity {SB j,1} m×1 is mapped to the job flow through the creating user of the job flow, forming a job flow similarity weighted coefficient matrix, denoted as {CA k,1} m×1 , where k represents the job flow serial number, and CA k,1 is equal to the similarity of the creating user of job flow k.

[0101] Step S23: Calculate the third similarity between the current job flow attribute and each second job flow attribute, and construct a job flow similarity matrix based on the third similarity. Determine the job flow weighted similarity matrix based on the job flow similarity matrix and the job flow similarity weighted coefficient matrix.

[0102] Optionally, in the technical solution provided in the above step S23, the second - type recommendation module can also determine the third similarity through the following method:

[0103] Step S231: Determine, through the second - type recommendation module, the number of identical job flow attributes that the current job flow and each second job flow have, and the total number of all job flow attributes that the current job flow and each second job flow have;

[0104] Step S232: Use the quotient of the number of identical job flow attributes and the total number of all job flow attributes as the third similarity between the current job flow attribute and each second job flow attribute.

[0105] In the above - mentioned embodiment, denote the current job flow attribute of the current job flow as A0, and denote the second job flow attributes of each of the other second job flows as A k , where k represents the job flow serial number of the other second job flows. Therefore, the number of identical job flow attributes that the current job flow and each second job flow have can be expressed as: |A0 ∩ A k |, and the total number of all job flow attributes that the current job flow and each second job flow have can be expressed as: |A0 ∪ A k |. Then, calculate the third similarity between the current job flow attribute and each second job flow attribute through the following formula:

[0106]

[0107] Finally, by repeating the above steps S231 - S232, a job flow similarity matrix between the current job flow and each of the other second job flows can be obtained, denoted as {SA k,1} m×1 .

[0108] Furthermore, by combining and calculating the job flow similarity matrix and the job flow similarity weighting coefficient matrix, a job flow weighted similarity matrix is obtained, and its expression is:

[0109] {SCA k,1} m×1 = {SA k,1} m×1 ⊙(1 + a1{CA k,1} m×1 )

[0110] where ⊙ represents the Hadamard product, a1 represents the weight of each job flow similarity weighting coefficient in the job flow similarity weighting coefficient matrix, and its value range is between 0 and 1. The larger this value is, the higher the importance of the user attribute. It should be noted that since the job flow attribute is closer to the atomic ability, and the user attribute indirectly affects the atomic ability through the job flow, in the embodiments of the present application, the weight of the job flow similarity weighting coefficient matrix is smaller.

[0111] Step S24, determine the job flow weighted similarity corresponding to the multiple second job flows where each atomic ability in the atomic ability library is located according to the job flow weighted similarity matrix, and determine the fourth similarity of the atomic ability according to the average value of the job flow weighted similarities corresponding to the multiple second job flows.

[0112] The above steps can be understood as follows: an atomic ability may exist in multiple job flows, so the fourth similarity of the atomic ability is the average value of the weighted similarities of the second job flows where it is located.

[0113] Step S25, construct an atomic ability similarity matrix based on the fourth similarity of each atomic ability.

[0114] Among them, the atomic ability similarity matrix can be constructed from the fourth similarity of each atomic ability, and its expression is: {S a,1}, where S m×1 is the fourth similarity of atomic ability a. a,1

[0115] ​Regarding the third type of recommendation module, the recommendation principle of the third type of recommendation module is as follows: During the job flow configuration process, the configured processes and atomic capabilities can find similar processes in the existing job flows, and such processes can be called similar sub-processes. By combining and calculating the similarity and scores of these similar sub-processes, the corresponding recommendation coefficient can be obtained. A high recommendation coefficient indicates that the recommended degree of the atomic capabilities corresponding to the downstream nodes of these sub-processes is also relatively high, thereby realizing the recommendation of atomic capabilities. The analysis process of the third type of recommendation module will be described below through the following steps S31 - S35:

[0116] Step S31: Extract the first upstream sub-process of the i-th working node with the first atomic capability configured in the current job flow through the third type of recommendation module, and the second upstream sub-processes of each working node with the second atomic capability configured in multiple other second job flows.

[0117] Specifically, the above step S31 can be understood as extracting the first atomic capability of the i-th working node in the current job flow and determining the first upstream sub-process of the i-th working node. Among them, the first upstream sub-process can be understood as a partial workflow composed of the first i working nodes configured in the current job flow; at the same time, extract the second atomic capabilities in the existing second job flows. Since each atomic capability may exist in multiple job flows and can appear on multiple working nodes in the same job flow, all the second upstream sub-processes of the nodes where each atomic capability is located are extracted. Thus, a mapping table of atomic capabilities and their upstream sub-processes can be formed, as shown in Table 2 below:

[0118] Table 2

[0119] Atomic ability Upstream subprocess Job flow where the upstream subprocess is located <![CDATA[a1]]> F1 Z1 <![CDATA[a2]]> F2 Z1 <![CDATA[a3]]> F3 Z2 <![CDATA[a4]]> F4 Z3 … … … <![CDATA[a m > <![CDATA[F l > <![CDATA[Z k >

[0120] Step S32: Construct multiple first adjacency matrices of the first upstream sub-process and multiple second upstream sub-processes respectively, and multiple second adjacency matrices of each second upstream sub-process and the first upstream sub-process.

[0121] Optionally, in the technical solution provided in the above step S32, this method can be implemented through the following steps S321 - S322:

[0122] Step S321: For each second upstream sub-process, divide the first upstream sub-process and the second upstream sub-process into the first type of atomic capabilities, the second type of atomic capabilities, and the third type of atomic capabilities through the third type of recommendation module. Among them, the first type of atomic capabilities is used to reflect the atomic capabilities unique to the first upstream sub-process, the second type of atomic capabilities is used to reflect the atomic capabilities shared by the first upstream sub-process and the second upstream sub-process, and the third type of atomic capabilities is used to reflect the atomic capabilities unique to the second upstream sub-process;

[0123] Step S322: Take the first type of atomic capabilities, the second type of atomic capabilities, and the third type of atomic capabilities as rows and columns respectively, and construct the first adjacency matrix of the first upstream sub-process and the second upstream sub-process, as well as the second adjacency matrix of the second upstream sub-process and the first upstream sub-process.

[0124] In the above embodiment, denote the first upstream sub-process as F0 and the second sub-process as F r , where F r = {F1, F2, …, F l}, r represents the serial number of the second sub-process, and l represents the number of the second sub-processes. Then, the atomic capabilities within the first sub-process and the atomic capabilities within the second sub-process can be divided into the following three categories: the first type of atomic capabilities unique to the first upstream sub-process F0, the second type of atomic capabilities shared by the first upstream sub-process F0 and the second upstream sub-process F r , and the third type of atomic capabilities unique to the second upstream sub-process F r .

[0125] Then, use these three types of atomic capabilities to construct an adjacency matrix for the first upstream sub-process F0 and the second upstream sub-process F r respectively, and arrange these three types of atomic capabilities continuously within the adjacency matrix. For example, the rows and columns within the adjacency matrix are set as follows: Y1 to Y h are the first type of atomic capabilities, set Y h+1 to Y g as the second type of atomic capabilities, and set Y g+1 to Y m as the third type of atomic capabilities. And the assignment rule for each element within the matrix is:

[0126] (1) If the atomic capability Y a exists, then assign the weight coefficient e of the existing atomic capability to W a,a ;

[0127] (2) If there is a directed edge connecting the atomic capabilities Y a and Y b , then a weight coefficient needs to be assigned to the directed edge. The rule is: when a < b, assign the weight coefficient d of the directed edge to W a,b , and assign 1 to W b,a ; when b < a, assign the weight coefficient d of the directed edge to W b,a , and assign 1 to W a,b .

[0128] It should be noted that the above W a,a , W a,b , W b,aBoth represent matrices, and the weight coefficient d of the directed edge should be greater than or equal to 1; when it is equal to 1, the algorithm will ignore the directionality of the directed edge. As long as there is a connection between atomic capabilities, it is considered exactly the same. The larger d is, the greater the difference between two sets of atomic capabilities with a connection but opposite directions.

[0129] Therefore, the first upstream sub-process F0 and the second upstream sub-process F above r The first adjacency matrix of can be denoted as A(F0,F r ), which represents a matrix jointly constructed by the atomic capabilities of the upstream sub-processes F0 and F r . By assigning values with F0, the matrix shown in Table 3 below can be obtained.

[0130] Table 3

[0131]

[0132]

[0133] The second upstream sub-process F r The second adjacency matrix with the first upstream sub-process F0 can be denoted as: A(F r ,F0), which represents a matrix jointly constructed by the atomic capabilities of the upstream sub-processes F0 and F r . By assigning values with F r , the matrix shown in Table 4 below can be obtained.

[0134] Table 4

[0135]

[0136] Repeat the above process until the multiple first adjacency matrices of the first upstream sub-process F0 and multiple second upstream sub-processes F r , and the multiple second adjacency matrices of multiple second upstream sub-processes F r and the first upstream sub-process F0 are determined.

[0137] Step S33: For each second upstream sub-process, calculate the fifth similarity between the first upstream sub-process and the second upstream sub-process based on the first adjacency matrix between the first upstream sub-process and the second upstream sub-process and the second adjacency matrix between the second upstream sub-process and the first upstream sub-process, and determine the first recommendation coefficient of the second upstream sub-process according to the fifth similarity.

[0138] Optionally, in the technical solution provided in the above step S33, the third type of recommendation model can determine the fifth similarity through the following steps S331 - S333, including:

[0139] Step S331: Calculate the first Hadamard product of the first upstream sub-process and the second upstream sub-process and the second Hadamard product of the second upstream sub-process and the first upstream sub-process, multiply the result by 2 to obtain a similarity numerator matrix, and then take the square root of the sum of all the first matrix elements in the similarity numerator matrix to obtain the similarity numerator;

[0140] Step S332: Calculate the second Hadamard product of the second adjacency matrix with itself, combine it with the first Hadamard product to obtain a similarity denominator matrix, and then take the square root of the sum of all the second matrix elements in the similarity denominator matrix to obtain the similarity denominator;

[0141] Step S333: Calculate the fifth similarity between the first upstream sub-process and the second upstream sub-process based on the similarity numerator and the similarity denominator.

[0142] For example, let the first adjacency matrix be A(F0, F r ), and the second adjacency matrix be A(F r , F0). First, calculate the Hadamard product of the first adjacency matrix A(F0, F r ) and the second adjacency matrix A(F r , F0), and then multiply the result by the scalar 2 to obtain a similarity numerator matrix, denoted as: 2(A(F0, F r ) ⊙ A(F r , F0)); then take the square root of the sum of all elements in the similarity numerator matrix 2(A(F0, F r ) ⊙ A(F r , F0)) to obtain the similarity numerator.

[0143] Then, calculate the second Hadamard product of the second adjacency matrix with itself (A(F r , F0) ⊙ A(F r , F0)), and then combine it with the first Hadamard product (A(F0, F r ) ⊙ A(F r , F0)) to obtain a similarity denominator matrix, denoted as: (A(F0, F r ) ⊙ A(F r , F0)) + (A(F r , F0) ⊙ A(F r , F0)), and then take the square root of the sum of all the second matrix elements in the similarity denominator matrix to obtain the similarity denominator.

[0144] Finally, calculate the fifth similarity between the first upstream sub-process F0 and the second upstream sub-process F r using the similarity numerator and the similarity denominator calculated above. The expression is:

[0145]

[0146] Repeat the above steps S331 - S333 for multiple second upstream sub - processes, and calculate the fifth similarity between the first upstream sub - process F0 and each of the second upstream sub - processes F1 to F l .

[0147] Furthermore, the third - type recommendation module can also determine the first recommendation coefficient of the second upstream sub - process according to the fifth similarity. The specific steps include:

[0148] First, determine the execution success rate of the second upstream sub - process through the third - type recommendation module. This success rate can be understood as the ratio of the number of times the second upstream sub - process is executed to the number of successful executions. The number of successful executions refers to the situation where when each work node in the second upstream sub - process does not report an error during the execution of the second job flow where the second upstream sub - process is located, it is regarded as a successful execution once.

[0149] Then, determine the first recommendation coefficient of the second upstream sub - process based on the product of the fifth similarity and the execution success rate. Therefore, the expression of the first recommendation coefficient can be written as:

[0150] First recommendation coefficient = (Fifth similarity between F0 and F r ) * Execution success rate of F r

[0151] Step S34: For any second atomic ability within the second upstream sub - process, determine the second recommendation coefficient of the second atomic ability according to the first recommendation coefficients of the multiple second upstream sub - processes where the second atomic ability is located.

[0152] In the technical solution provided in the above step S34, according to Table 2 above, it can be known that the recommendation coefficient of a certain atomic ability can be calculated using the recommendation coefficients of all upstream sub - processes corresponding to the atomic ability. Therefore, in the embodiments of the present application, for any second atomic ability within the second upstream sub - process, the second recommendation coefficient of the second atomic ability can be determined using the first recommendation coefficients of the multiple second upstream sub - processes where the second atomic ability is located.

[0153] Specifically, the method for determining the second recommendation coefficient includes: taking the square root of the sum of the squares of the first recommendation coefficients of the multiple second upstream sub - processes where the second atomic ability is located as the second recommendation coefficient of the second atomic ability. Therefore, the second recommendation coefficient of the second atomic ability can be written as:

[0154]

[0155] Among them, F a represents all second upstream sub - processes where the second atomic ability a is located.

[0156] ​Step S35: A second recommendation coefficient matrix is formed by the second recommendation coefficients corresponding to each second atomic ability.

[0157] Specifically, the above steps are executed for all atomic abilities in the mapping table shown in Table 2 to obtain the second recommendation coefficients of all atomic abilities, and a second recommendation coefficient matrix is formed by these second recommendation coefficients, denoted as {T a,1} m×1 .

[0158] Combining the first type of recommendation module, the second type of recommendation module, and the third type of recommendation module introduced above, the atomic recommendation process proposed in steps S206 and S208 is divided into Figure 3 and Figure 4 two cases, where:

[0159] As Figure 3 shown, when the number of first atomic abilities is less than the preset quantity threshold, the process orchestration system can call the second type of recommendation module to analyze the first user attributes and the first job flow attributes, and recommend a first batch of atomic abilities to the first user, enabling the first user to select a target atomic ability from the first batch of atomic abilities as the first working node; subsequently, the process orchestration system can also call the hybrid recommendation module 2 composed of the second type of recommendation module and the third type of recommendation module to recommend a batch of atomic abilities for the (i + 1)-th working node based on the first user attributes, the first job flow attributes, and the target atomic abilities of the previous i working nodes, enabling the first user to select a target atomic ability from this batch of atomic abilities, and repeating this step until the target atomic abilities are recommended for all working nodes of the first workflow.

[0160] As Figure 4 shown, when the number of first atomic abilities is not less than the preset quantity threshold, the process orchestration system can call the hybrid recommendation module 1 composed of the first type of recommendation module and the second type of recommendation module to analyze the first user attributes, the first job flow attributes, and the first scores of the first user for each atomic ability, and recommend a first batch of atomic abilities to the first user, enabling the first user to select a target atomic ability from the first batch of atomic abilities; subsequently, the process orchestration system can also call the hybrid recommendation module 3 composed of the first type of recommendation module, the second type of recommendation module, and the third type of recommendation module to recommend a batch of atomic abilities for the (i + 1)-th working node based on the first user attributes, the first job flow attributes, the first scores of the first user for each atomic ability, and the target atomic abilities of the previous i working nodes, enabling the first user to select a target atomic ability from this batch of atomic abilities, and repeating this step until the target atomic abilities are recommended for all working nodes of the first workflow.

[0161] Therefore, in the technical solution provided in the above step S206, the method may include the following steps S2061 - S2062:

[0162] Step S2061: Invoke the second - type recommendation module to analyze the first user attributes and the first job - flow attributes, obtain the first atomic - ability similarity matrix, and recommend the target atomic ability of the first job node of the first job - flow to the first user according to the first atomic - ability similarity matrix.

[0163] Specifically, the process - orchestration system analyzes the first user attributes and the first job - flow attributes by invoking the second - type recommendation module, obtains the first atomic - ability similarity matrix, denoted as {S a,1} m×1 , where each element in the first atomic - ability similarity matrix is the first atomic - ability similarity. Thus, at least one target atomic ability of the first job node can be recommended to the first user in descending order of the first atomic - ability similarity, enabling the first user to select one target atomic ability from at least one target atomic ability.

[0164] Step S2062: Invoke the second - type recommendation module and the third - type recommendation module to analyze the first user attributes, the first job - flow attributes, and the target atomic abilities of the first i job nodes, obtain the first atomic - ability similarity matrix and the second target - recommendation coefficient matrix, and recommend the target atomic ability corresponding to the (i + 1)-th job node of the first job - flow to the first user according to the first atomic - ability similarity matrix and the second target - recommendation coefficient matrix.

[0165] Optionally, in the technical solution provided in the above step S2062, the method further includes: recommending the target atomic ability corresponding to the (i + 1)-th job node of the first job - flow to the first user according to the first atomic - ability similarity matrix, the weight factor of the first atomic - ability similarity matrix, and the second target - recommendation coefficient matrix.

[0166] Specifically, the process - orchestration system analyzes the first user attributes and the first job - flow attributes by invoking the second - type recommendation module, obtains the first atomic - ability similarity matrix, denoted as {S a,1} m×1 , and analyzes the target atomic abilities of the first i job nodes by invoking the third - type recommendation module, obtains the second target - recommendation coefficient matrix, denoted as {T a,1} m×1 , and then determines the recommendation result through the following formula:

[0167] Recommendation result={T a,1} m×1 ⊙(1 + c{S a,1} m×1 )

[0168] Among them, c represents the weight factor of the first atomic ability similarity matrix, and the value range of c is between 0 and 1. The larger it is, the higher the importance of the atomic ability similarity calculated by the second type of recommendation module.

[0169] For example, if the process orchestration system needs to recommend a target atomic ability for the second work node of the first workflow created by the first user, the process orchestration system can call the second type of recommendation module to analyze the first user attributes and the first job flow attributes to obtain the first atomic ability similarity matrix, denoted as {S a,1} m×1 and call the third type of recommendation module to analyze the target atomic ability of the first job node, and then recommend the most suitable target atomic ability to the second work node. If the process orchestration system needs to recommend a target atomic ability for the eighth work node of the first workflow created by the first user, the process orchestration system can call the second type of recommendation module to analyze the first user attributes and the first job flow attributes to obtain the first atomic ability similarity matrix, denoted as {S a,1} m×1 and call the third type of recommendation module to analyze the target atomic ability of the seventh job node, and then recommend the most suitable target atomic ability to the eighth work node.

[0170] In the technical solution provided in step S208 above, the method may include the following steps S2081-S2082:

[0171] Step S2081: Call the first type of recommendation module and the second type of recommendation module to analyze the first user attributes, the first job flow attributes, and the first user's scores for each atomic ability, obtain the first target recommendation coefficient matrix and the first atomic ability similarity matrix, and recommend the target atomic ability of the first job node of the first job flow to the first user based on the first target recommendation coefficient matrix and the first atomic ability similarity matrix.

[0172] Optionally, in the technical solution provided in step S2081 above, the method further includes: recommending the target atomic ability of the first job node of the first job flow to the first user based on the first atomic ability similarity matrix, the weight factor of the first atomic ability similarity matrix, and the first target recommendation coefficient matrix.

[0173] Specifically, the process orchestration system analyzes the first scores of the first user for each atomic ability by calling the first type of recommendation module to obtain the first target recommendation coefficient matrix, denoted as {P a,1} m×1 and calls the second type of recommendation module to analyze the first user attributes and the first job flow attributes to obtain the first atomic ability similarity matrix {S a,1} m×1, and then determine the recommendation result through the following formula:

[0174] Recommendation result = {P a,1} m×1 ⊙(1 + c{S a,1} m×1 )

[0175] Step S2082: Invoke the first - type recommendation module, the second - type recommendation module, and the third - type recommendation module to analyze the first user attributes, the first job - flow attributes, the scores of the first user for each atomic ability, and the target atomic abilities of the first i job nodes, obtain the first target recommendation coefficient matrix, the first atomic - ability similarity matrix, and the second target recommendation coefficient matrix, and recommend the target atomic ability corresponding to the (i + 1)-th job node of the first job - flow to the first user based on the first target recommendation coefficient matrix, the first atomic - ability similarity matrix, and the second target recommendation coefficient matrix.

[0176] Optionally, in the technical solution provided in the above step S2082, the method further includes: recommending the target atomic ability corresponding to the (i + 1)-th job node of the first job - flow to the first user based on the first atomic - ability similarity matrix, the weight factor of the first atomic - ability similarity matrix, the first target recommendation coefficient matrix, and the second target recommendation coefficient matrix.

[0177] Specifically, the process - orchestration system analyzes the first scores of the first user for each atomic ability by invoking the first - type recommendation module to obtain the first target recommendation coefficient matrix, denoted as {P a,1} m×1 , and analyzes the first user attributes and the first job - flow attributes by invoking the second - type recommendation module to obtain the first atomic - ability similarity matrix {S a,1} m×1 , and analyzes the target atomic abilities of the first i job nodes by invoking the third - type recommendation module to obtain the second target recommendation coefficient matrix {T a,1} m×1 , and then determine the recommendation result through the following formula:

[0178] Recommendation result = ({P a,1} m×1 +{T a,1} m×1 )⊙(1 + c{S a,1} m×1 )

[0179] In the above steps, by analyzing the user's historical behavior, the frequently used atomic capabilities of the user can be recommended, and the atomic capabilities that the user has not used but may be useful can also be recommended; by analyzing the job flow attributes of the current job flow and the user attributes of the current user, atomic capabilities can be recommended for the user, and at the same time, the recommendation can be made more accurate; by identifying the topological structure of the job flow and the configured atomic capabilities, excellent atomic capabilities that may be used can be found from the existing job flows; by accurately recommending the atomic capabilities of each node in real time during the configuration process of the job flow, the development efficiency and development quality can be effectively improved. The embodiments of the present application can be widely applied to the atomic capability recommendation in the orchestration scenario under various scenarios, and the recommendation accuracy of the atomic capabilities is higher.

[0180] Embodiment 2

[0181] Based on Embodiment 1 of the present application, an embodiment of an atomic capability recommendation device for process orchestration is further provided. When the device runs, it executes the above-mentioned atomic capability recommendation method for process orchestration in the above-mentioned embodiment. Among them, Figure 5 is a schematic structural diagram of an optional atomic capability recommendation device for process orchestration according to an embodiment of the present application, as Figure 5 shown. The atomic capability recommendation device for process orchestration includes at least a determination module 51, a comparison module 52, a first recommendation module 53, and a second recommendation module 54, where:

[0182] The determination module 51 is configured to obtain a first job flow to be orchestrated by a first user, and determine the first user attribute, the first historical scoring behavior data of the first user, and the first job flow attribute of the first job flow. Among them, the first historical scoring behavior data includes: the first score of the first user for each atomic capability in the atomic capability library, and the number of first atomic capabilities scored by the first user.

[0183] The comparison module 52 is configured to compare the number of first atomic capabilities scored by the first user with a preset quantity threshold.

[0184] The first recommendation module 53 is configured to, when the number of first atomic capabilities is less than the preset quantity threshold, recommend multiple target atomic capabilities for orchestrating the first job flow to the first user according to the first user attribute and the first job flow attribute.

[0185] The second recommendation module 54 is configured to, when the number of atomic capabilities is not less than the preset quantity threshold, recommend multiple target atomic capabilities for orchestrating the first job flow to the first user according to the first user attribute, the first job flow attribute, and the first score of the first user for each atomic capability.

[0186] It should be noted that each module in the above atomic capability recommendation device for process orchestration can be a program module (for example, a set of program instructions that implement a specific function), or a hardware module. For the latter, it can be presented in the following forms, but not limited to: the manifestation of each of the above modules is a processor, or the functions of each of the above modules are implemented by a processor.

[0187] Embodiment 3

[0188] According to an embodiment of the present application, a non-volatile storage medium is further provided. A program is stored in the non-volatile storage medium. When the program runs, it controls the device where the non-volatile storage medium is located to execute the atomic capability recommendation method for process orchestration in Embodiment 1.

[0189] Optionally, the device where the non-volatile storage medium is located executes the following steps by running the program:

[0190] Step S202: Obtain the first job flow to be orchestrated by the first user, and determine the first user attribute, the first historical scoring behavior data of the first user, and the first job flow attribute of the first job flow. Among them, the first historical scoring behavior data includes: the first scores of the first user for each atomic capability in the atomic capability library, and the number of the first atomic capabilities scored by the first user.

[0191] Step S204: Compare the number of the first atomic capabilities scored by the first user with a preset quantity threshold.

[0192] Step S206: When the number of the first atomic capabilities is less than the preset quantity threshold, recommend multiple target atomic capabilities for orchestrating the first job flow to the first user according to the first user attribute and the first job flow attribute.

[0193] Step S208: When the number of atomic capabilities is not less than the preset quantity threshold, recommend multiple target atomic capabilities for orchestrating the first job flow to the first user according to the first user attribute, the first job flow attribute, and the first scores of the first user for each atomic capability.

[0194] The serial numbers of the above embodiments of the present application are only for description and do not represent the advantages and disadvantages of the embodiments.

[0195] In the above embodiments of the present application, the descriptions of each embodiment have their own focuses. For the parts not detailed in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0196] In several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are merely illustrative. For example, the division of the units can be a logical function division. In actual implementation, there can be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection between each other can be through some interfaces. The indirect coupling or communication connection of units or modules can be in an electrical or other form.

[0197] The units described as separate components may or may not be physically separated. The components displayed as units may or may not be physical units, that is, they can be located in one place or distributed to multiple units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0198] In addition, in each embodiment of this application, each functional unit can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit. The above-mentioned integrated units can be implemented in the form of hardware or in the form of software functional units.

[0199] If the above-mentioned integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of this application, in essence, or the part that contributes to the relevant technology, or all or part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in each embodiment of this application. The foregoing storage medium includes: various media such as USB flash drives, read-only memories (ROMs), random access memories (RAMs), mobile hard disks, magnetic disks, or optical discs that can store program codes.

[0200] The above is only the preferred embodiment of this application. It should be noted that for those of ordinary skill in the art, without departing from the principle of this application, several improvements and refinements can still be made, and these improvements and refinements should also be regarded as the protection scope of this application.

Claims

1. An atomic capability recommendation method for process orchestration, characterized in that, Including: Obtain a first job flow to be orchestrated by a first user, and determine a first user attribute, first historical scoring behavior data of the first user, and a first job flow attribute of the first job flow. Wherein, the first historical scoring behavior data includes: a first score of the first user for each atomic ability in the atomic ability library, and the number of first atomic abilities scored by the first user; Compare the number of first atomic abilities scored by the first user with a preset quantity threshold; When the number of first atomic abilities is less than the preset quantity threshold, recommend multiple target atomic abilities for orchestrating the first job flow to the first user according to the first user attribute and the first job flow attribute; When the number of atomic abilities is not less than the preset quantity threshold, recommend multiple target atomic abilities for orchestrating the first job flow to the first user according to the first user attribute, the first job flow attribute, and the first score of the first user for each atomic ability.

2. The method according to claim 1, wherein Obtain a first job flow to be orchestrated by a first user, and determine a first user attribute, first historical scoring behavior data, and a first job flow attribute of the first job flow, including: Obtain the first job flow to be orchestrated by the first user in the process orchestration system, and determine a first user attribute of the first user and a first job flow attribute of the first job flow. Wherein, the first user attribute includes at least one of the following: user role, user department, user position, and the first job flow attribute includes at least one of the following: process purpose, process group, process label; For each atomic ability in the atomic ability library, obtain the scoring behavior of the first user for the atomic ability, and determine a first score of the first user for the atomic ability according to the scoring behavior and a preset behavior weight. Wherein, the scoring behavior includes at least one of the following: the first user favorites the atomic ability, any job flow orchestrated by the first user uses the atomic ability, the first user views the atomic ability; Summarize the first scores of the first user for each atomic ability, and determine the number of first atomic abilities scored by the first user.

3. The method according to claim 1, wherein The method for determining the preset quantity threshold includes: Determine the number of atomic abilities scored by multiple users in the process orchestration system, and determine the median and average of the array composed of the number of atomic abilities scored by each user; Calculate the average of the median and the average, and use the average as the preset quantity threshold.

4. The method according to claim 1, wherein The method further includes: Create a first type of recommendation module, where the first type of recommendation module is used to analyze the historical scoring behavior data of each user in the process orchestration system to obtain a first recommendation coefficient matrix of each user for each atomic ability in the atomic ability library; Create a second type of recommendation module, where the second type of recommendation module is used to analyze the current job flow attributes of the current job flow to be choreographed by the current user within the process choreography system, the current user attributes, the second user attributes of multiple other second users, and the second job flow attributes of multiple other second job flows, to obtain the atomic ability similarity matrix of each atomic ability in the atomic ability library; Create a third type of recommendation module, where the third type of recommendation module is used to analyze the first atomic ability of the first i working nodes in the current job flow, to obtain the second recommendation coefficient matrix of each atomic ability in the atomic ability library, where i is a positive integer greater than 1.

5. The method according to claim 4, characterized in that The analysis process of the first type of recommendation module includes: Determine the second score of each user for each atomic ability in the atomic ability library through the first type of recommendation module; Construct a score matrix based on the second scores of the users for each atomic ability in the atomic ability library, and calculate the first similarity of any two atomic abilities based on the second scores of the users for each atomic ability in the atomic ability library, and construct the atomic ability similarity matrix from the first similarity of any two atomic abilities; Multiply the score matrix by the atomic ability similarity matrix to obtain the first recommendation coefficient matrix of each user for each atomic ability in the atomic ability library.

6. The method according to claim 4, characterized in that The analysis process of the second type of recommendation module includes: Calculate the second similarity between the current user attributes of the current user and the second user attributes of each second user through the second type of recommendation module, and construct a user similarity matrix based on the second similarity; Convert the user similarity matrix into a job flow similarity weighted coefficient matrix for characterizing the similarity of the creating users of the job flow; Calculate the third similarity between the current job flow attributes and each second job flow attribute, and construct a job flow similarity matrix based on the third similarity, and determine the job flow weighted similarity matrix based on the job flow similarity matrix and the job flow similarity weighted coefficient matrix; Determine the job flow weighted similarity corresponding to the multiple second job flows where each atomic ability in the atomic ability library is located based on the job flow weighted similarity matrix, and determine the fourth similarity of the atomic ability based on the average value of the job flow weighted similarities corresponding to the multiple second job flows; Construct the atomic ability similarity matrix based on the fourth similarity of each atomic ability.

7. The method according to claim 6, wherein Calculating the second similarity between the current user attributes of the current user and the second user attributes of each second user through the second type of recommendation module includes: Determine the number of identical user attributes possessed by the current user and each second user, and the total number of user attributes possessed by the current user and each second user through the second type of recommendation module; Use the quotient of the number of identical user attributes and the total number of user attributes as the second similarity between the current user attributes and the second user attributes of each second user.

8. The method according to claim 6, characterized in that, Calculating the third similarity between the current job flow attribute and each of the second job flow attributes includes: Determining, by the second type of recommendation module, the number of identical job flow attributes between the current job flow and each of the second job flows, and the total number of job flow attributes of the current job flow and each of the second job flows; Taking the quotient of the number of identical job flow attributes and the total number of job flow attributes as the third similarity between the current job flow attribute and each of the second job flow attributes.

9. The method according to claim 4, characterized in that The analysis process of the third type of recommendation module includes: Extracting, by the third type of recommendation module, the first upstream sub-process of the i-th working node with the first atomic capability configured in the current job flow and the second upstream sub-processes of each working node with the second atomic capability configured in multiple other second job flows; Constructing multiple first adjacency matrices between the first upstream sub-process and the multiple second upstream sub-processes, and multiple second adjacency matrices between each of the second upstream sub-processes and the first upstream sub-process; For each of the second upstream sub-processes, calculating the fifth similarity between the first upstream sub-process and the second upstream sub-process based on the first adjacency matrix between the first upstream sub-process and the second upstream sub-process and the second adjacency matrix between the second upstream sub-process and the first upstream sub-process, and determining the first recommendation coefficient of the second upstream sub-process according to the fifth similarity; For any second atomic capability in the second upstream sub-process, determining the second recommendation coefficient of the second atomic capability according to the first recommendation coefficients of the multiple second upstream sub-processes where the second atomic capability is located; Forming a second recommendation coefficient matrix from the second recommendation coefficients corresponding to each of the second atomic capabilities.

10. The method according to claim 9, wherein Constructing multiple first adjacency matrices between the first upstream sub-process and the multiple second upstream sub-processes, and multiple second adjacency matrices between each of the second upstream sub-processes and the first upstream sub-process, including: For each of the second upstream sub-processes, dividing, by the third type of recommendation module, the first upstream sub-process and the second upstream sub-process into a first type of atomic capability, a second type of atomic capability, and a third type of atomic capability, where the first type of atomic capability is used to reflect the unique atomic capability of the first upstream sub-process, the second type of atomic capability is used to reflect the common atomic capability between the first upstream sub-process and the second upstream sub-process, and the third type of atomic capability is used to reflect the unique atomic capability of the second upstream sub-process; Taking the first type of atomic capability, the second type of atomic capability, and the third type of atomic capability as rows and columns respectively, and constructing the first adjacency matrix between the first upstream sub-process and the second upstream sub-process, and the second adjacency matrix between the second upstream sub-process and the first upstream sub-process respectively.

11. The method according to claim 9, wherein Calculating the fifth similarity between the first upstream sub-process and the second upstream sub-process based on the first adjacency matrix between the first upstream sub-process and the second upstream sub-process and the second adjacency matrix between the second upstream sub-process and the first upstream sub-process, including: Calculate the first Hadamard product of the first adjacency matrix of the first upstream sub-process and the second upstream sub-process and the second adjacency matrix of the second upstream sub-process and the first upstream sub-process by the third type of recommendation module, multiply the result by 2 to obtain a similarity numerator matrix, and take the square root after summing up each first matrix element in the similarity numerator matrix to obtain a similarity numerator; Calculate the second Hadamard product of the second adjacency matrix itself, combine it with the first Hadamard product to obtain a similarity denominator matrix, and take the square root after summing up each second matrix element in the similarity denominator matrix to obtain a similarity denominator; Calculate the fifth similarity between the first upstream sub-process and the second upstream sub-process based on the similarity numerator and the similarity denominator.

12. The method according to claim 9, wherein Calculate the first recommendation coefficient of the second upstream sub-process according to the fifth similarity, including: Determine the execution success rate of the second upstream sub-process by the third type of recommendation module; Determine the first recommendation coefficient of the second upstream sub-process based on the product of the fifth similarity and the execution success rate.

13. The method according to claim 9, characterized in that, Determine the second recommendation coefficient of the second atomic ability according to the first recommendation coefficients of the multiple second upstream sub-processes where the second atomic ability is located, including: Take the result obtained by taking the square root of the sum of squares of the first recommendation coefficients of the multiple second upstream sub-processes where the second atomic ability is located by the third type of recommendation module as the second recommendation coefficient of the second atomic ability.

14. The method according to claim 4, wherein Recommend multiple target atomic abilities for choreographing the first job flow to the first user according to the first user attribute and the first job flow attribute, including: Call the second type of recommendation module to analyze the first user attribute and the first job flow attribute to obtain a first atomic ability similarity matrix, and recommend the target atomic ability of the first job node of the first job flow to the first user according to the first atomic ability similarity matrix; Call the second type of recommendation module and the third type of recommendation module to analyze the first user attribute, the first job flow attribute and the target atomic abilities of the first i job nodes to obtain a first atomic ability similarity matrix and a second target recommendation coefficient matrix, and recommend the target atomic ability corresponding to the (i + 1)-th job node of the first job flow to the first user according to the first atomic ability similarity matrix and the second target recommendation coefficient matrix.

15. The method according to claim 14, wherein Recommend the target atomic ability corresponding to the (i + 1)-th job node of the first job flow to the first user according to the first atomic ability similarity matrix and the second target recommendation coefficient matrix, including: Recommend the target atomic ability corresponding to the (i + 1)-th job node of the first job flow to the first user according to the first atomic ability similarity matrix, the weight factor of the first atomic ability similarity matrix, and the second target recommendation coefficient matrix.

16. The method according to claim 4, characterized in that, Recommend multiple target atomic abilities for choreographing the first job flow to the first user according to the first user attribute, the first job flow attribute, and the scores of the first user for each atomic ability, including: Call the first type of recommendation module and the second type of recommendation module to analyze the first user attributes, the first job flow attributes, and the first user's first scores for each of the atomic capabilities, obtain a first target recommendation coefficient matrix and a first atomic capability similarity matrix, and recommend the target atomic capability of the first job node of the first job flow to the first user based on the first target recommendation coefficient matrix and the first atomic capability similarity matrix; Call the first type of recommendation module, the second type of recommendation module, and the third type of recommendation module to analyze the first user attributes, the first job flow attributes, the first user's first scores for each of the atomic capabilities, and the target atomic capabilities of the first i job nodes, obtain a first target recommendation coefficient matrix, a first atomic capability similarity matrix, and a second target recommendation coefficient matrix, and recommend the target atomic capability corresponding to the (i + 1)-th job node of the first job flow to the first user based on the first target recommendation coefficient matrix, the first atomic capability similarity matrix, and the second target recommendation coefficient matrix.

17. The method according to claim 16, wherein Recommending the target atomic capability of the first job node of the first job flow to the first user based on the first target recommendation coefficient matrix and the first atomic capability similarity matrix includes: Recommending the target atomic capability of the first job node of the first job flow to the first user based on the first atomic capability similarity matrix, the weight factor of the first atomic capability similarity matrix, and the first target recommendation coefficient matrix.

18. The method according to claim 16, wherein Recommending the target atomic capability corresponding to the (i + 1)-th job node of the first job flow to the first user based on the first target recommendation coefficient matrix, the first atomic capability similarity matrix, and the second target recommendation coefficient matrix includes: Recommending the target atomic capability corresponding to the (i + 1)-th job node of the first job flow to the first user based on the first atomic capability similarity matrix, the weight factor of the first atomic capability similarity matrix, the first target recommendation coefficient matrix, and the second target recommendation coefficient matrix.

19. An atomic capability recommendation device for process orchestration, characterized in that, including: A determination module, configured to obtain a first job flow to be orchestrated by a first user, and determine the first user attributes of the first user, the first historical scoring behavior data, and the first job flow attributes of the first job flow, where the first historical scoring behavior data includes: the first scores of the first user for each atomic capability in the atomic capability library, and the number of first atomic capabilities scored by the first user; A comparison module, configured to compare the number of first atomic capabilities scored by the first user with a preset quantity threshold; A first recommendation module, configured to, when the number of first atomic capabilities is less than the preset quantity threshold, recommend a plurality of target atomic capabilities for orchestrating the first job flow to the first user based on the first user attributes and the first job flow attributes; A second recommendation module, configured to recommend, to the first user, a plurality of target atomic capabilities for choreographing the first job flow according to the first user attribute, the first job flow attribute, and the first ratings of the first user for the respective atomic capabilities when the number of the atomic capabilities is not less than the preset quantity threshold.

20. A non-volatile storage medium, characterized in that, A computer program is stored in the non-volatile storage medium, wherein the device where the non-volatile storage medium is located executes the atomic capability recommendation method for process choreography according to any one of claims 1 to 18 by running the computer program.

Images