Method, device, terminal and medium for displaying federated learning workflow interface
By generating and adjusting visual interfaces in the federated learning workflow, filtering and changing display parameters according to role types, the problem that traditional display methods cannot protect data privacy and security is solved, and a secure display of the federated learning process involving multiple parties is achieved.
Patent Information
- Application Number
- CN201911057310.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2019-10-31
- Publication Date
- 2024-11-01
- Estimated Expiration
- 2039-10-31
AI Technical Summary
Traditional machine learning workflow presentation methods cannot effectively protect the data privacy and data security of federated learning participants, especially in the multi-participated federated learning process.
By generating a visual interface when the federated learning workflow is initiated, and adjusting the output display status of the interface according to the role type of the output object, filtering and changing the display parameters of the target work component to ensure that each participant can only see its own machine Learning process to prevent data leakage.
This realizes the privacy and security protection of each participant's data during the display of federated learning workflow, ensuring that each participant's machine learning process will not be known by other participants, and ensuring data privacy and security.
Smart Images

Figure CN110825476B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of Fintech (financial technology), and in particular to a method, device, terminal device and computer-readable storage medium for displaying a federated learning workflow interface. Background Art
[0002] In the field of machine learning, in order to make the entire operation process of machine learning clear and to detect machine learning anomalies in a timely manner, visualizing the machine learning workflow has gradually become a research direction that the industry pays more and more attention to.
[0003] However, in the process of federated machine learning, since federated learning involves the participation of multiple parties and each participant participates in a different stage of federated learning, for the sake of data privacy and data security, it is necessary to ensure that the machine learning process of each participant will not be known to other participants. In this way, the use of traditional machine learning workflow presentation methods will not be able to achieve the purpose of protecting the data privacy and data security of federated learning participants. Summary of the invention
[0004] The main purpose of the present invention is to provide a method, apparatus, terminal device and computer-readable storage medium for displaying a federated learning workflow interface, aiming to solve the technical problem that the traditional machine learning workflow display method cannot protect the data privacy and data security of federated learning participants.
[0005] To achieve the above object, the present invention provides a method for displaying a federated learning workflow interface, the method for displaying a federated learning workflow interface comprising:
[0006] When a federated learning workflow is initiated, generating a visualization interface of the federated learning workflow;
[0007] Detecting an output object of the visualization interface, and adjusting an output display state of the visualization interface according to the output object;
[0008] The output of the visualization interface after adjusting the output display state is displayed to the output object.
[0009] Furthermore, the step of adjusting the output display state of the visualization interface according to the output object includes:
[0010] Detecting the role type of the display output object in federated learning;
[0011] The output display state of the visualization interface is adjusted according to the role type.
[0012] Furthermore, the step of adjusting the output display state of the visualization interface according to the role type includes:
[0013] According to the role type, filter the target working components in the visualization interface that are invisible to the output object;
[0014] In the visualization interface, display parameters of the target working component are changed to adjust the output display state of the visualization interface.
[0015] Furthermore, before the step of generating a visualization interface of the federated learning workflow when the federated learning workflow is initiated, the method further includes:
[0016] Configure each running component in the federated learning workflow based on the preset federated learning task,
[0017] The step of generating a visualization interface of the federated learning workflow includes:
[0018] The visualization interface is generated accordingly according to the operation process of each operation component and the display parameters.
[0019] Furthermore, the step of configuring each running component in the federated learning workflow based on the preset federated learning task includes:
[0020] Read the workflow configuration information carried by the preset federated learning task;
[0021] According to the workflow configuration information, configure an independently running work component in the federated learning workflow and specify a unilateral execution role of the independently running work component, and / or configure a jointly running work component in the federated learning workflow and specify a multi-party execution role of the jointly running work component;
[0022] According to the workflow configuration information, the operating parameter information of the independently running working component and / or the jointly running working component is configured in the federated learning workflow.
[0023] Furthermore, after the step of outputting and displaying the visual interface after adjusting the output display state to the output object, the method further includes:
[0024] A request instruction for viewing component output results is obtained, and the output results of the running component specified by the request instruction are displayed to the target output object.
[0025] Furthermore, the step of displaying the output result of the running component specified by the request instruction to the target output object includes:
[0026] Obtaining the operator of the operating component specified by the request instruction, and detecting whether the triggering end of the request instruction belongs to the operator;
[0027] If so, the trigger end is used as the target output object to output and display the output result.
[0028] In addition, to achieve the above-mentioned purpose, the present invention further provides a display device for a federated learning workflow interface, wherein the display device for the federated learning workflow interface comprises:
[0029] A generation module, used to generate a visual interface of the federated learning workflow when the federated learning workflow is initiated;
[0030] An adjustment module, used for detecting an output object of the visualization interface, and adjusting an output display state of the visualization interface according to the output object;
[0031] The display module is used to display the output of the visual interface after adjusting the output display state to the output object.
[0032] The present invention also provides a terminal device, comprising: a memory, a processor, and a display program for a federated learning workflow interface stored in the memory and executable on the processor, wherein the display program for the federated learning workflow interface implements the steps of the method for displaying the federated learning workflow interface as described above when executed by the processor.
[0033] The present invention also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the method for displaying a federated learning workflow interface as described above are implemented.
[0034] The method, device, terminal device and computer-readable storage medium for displaying the federated learning workflow interface proposed in the present invention generate a visualization interface of the federated learning workflow when the federated learning workflow is initiated; detect the output object of the visualization interface, adjust the output display state of the visualization interface according to the output object; and output the visualization interface after adjusting the output display state to the output object. The present invention adjusts the output display state of the visualization interface according to the different output objects of the current visualization interface in the federated learning when outputting and displaying the visualization interface of the federated learning workflow, and outputs and displays the visualization interface of the adjusted federated learning workflow to the current display object, so that when displaying the federated learning workflow involving multiple parties, different output display states of the visualization interface are displayed for different participants, ensuring that the machine learning process of each participant will not be known to other participants, thereby achieving the purpose of protecting the data privacy and data security of each participant in the federated learning in the process of displaying the federated learning workflow. BRIEF DESCRIPTION OF THE DRAWINGS
[0035] Figure 1It is a schematic diagram of the structure of the hardware operation involved in the embodiment of the present invention;
[0036] Figure 2 It is a flowchart of a first embodiment of a method for displaying a federated learning workflow interface of the present invention;
[0037] Figure 3A It is a schematic diagram of an interface output display scene of an embodiment of a method for displaying a federated learning workflow interface of the present invention;
[0038] Figure 3B It is another interface output display scene schematic diagram of an embodiment of a method for displaying a federated learning workflow interface of the present invention;
[0039] Figure 3C It is another interface output display scene schematic diagram of an embodiment of a method for displaying a federated learning workflow interface of the present invention;
[0040] Figure 4 It is a structural schematic diagram of a display device of a federated learning workflow interface of the present invention.
[0041] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION
[0042] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.
[0043] like Figure 1 As shown, Figure 1 It is a structural diagram of the hardware operating environment involved in the embodiment of the present invention.
[0044] It should be noted that Figure 1 That is, it is a schematic diagram of the structure of the hardware operating environment of the terminal device. The terminal device in the embodiment of the present invention can be a PC, a portable computer and other terminal devices.
[0045] like Figure 1As shown, the terminal device may include: a processor 1001, such as a CPU, a network interface 1004, a user interface 1003, a memory 1005, and a communication bus 1002. Among them, the communication bus 1002 is used to realize the connection and communication between these components. The user interface 1003 may include a display screen (Display), an input unit such as a keyboard (Keyboard), and the optional user interface 1003 may also include a standard wired interface and a wireless interface. The network interface 1004 may optionally include a standard wired interface and a wireless interface (such as a WI-FI interface). The memory 1005 may be a high-speed RAM memory, or a stable memory (non-volatile memory), such as a disk memory. The memory 1005 may also be a storage device independent of the aforementioned processor 1001.
[0046] Those skilled in the art will understand that Figure 1 The terminal device structure shown in the figure does not constitute a limitation on the terminal device, and may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently.
[0047] like Figure 1 As shown, the memory 1005 as a computer storage medium may include an operating system, a network communication module, a user interface module, and a distributed task processing program. The operating system is a program that manages and controls the hardware and software resources of the sample terminal device, and supports the operation of the distributed task processing program and other software or programs.
[0048] exist Figure 1 In the terminal device shown, the user interface 1003 is mainly used to communicate data with each terminal; the network interface 1004 is mainly used to connect to the background server and communicate data with the background server; and the processor 1001 can be used to call the display program of the federated learning workflow interface stored in the memory 1005, and perform the following operations:
[0049] When a federated learning workflow is initiated, generating a visualization interface of the federated learning workflow;
[0050] Detecting an output object of the visualization interface, and adjusting an output display state of the visualization interface according to the output object;
[0051] The output of the visualization interface after adjusting the output display state is displayed to the output object.
[0052] Further, the processor 1001 may call a display program of the federated learning workflow interface stored in the memory 1005, and further perform the following operations:
[0053] Detecting the role type of the display output object in federated learning;
[0054] The output display state of the visualization interface is adjusted according to the role type.
[0055] Further, the processor 1001 may call a display program of the federated learning workflow interface stored in the memory 1005, and further perform the following operations:
[0056] According to the role type, filter the target working components in the visualization interface that are invisible to the output object;
[0057] In the visualization interface, display parameters of the target working component are changed to adjust the output display state of the visualization interface.
[0058] Further, the processor 1001 may call a display program of the federated learning workflow interface stored in the memory 1005, and before generating a visualization interface of the federated learning workflow when the federated learning workflow is initiated, further perform the following operations:
[0059] Each running component in the federated learning workflow is configured based on a preset federated learning task.
[0060] Further, the processor 1001 may call a display program of the federated learning workflow interface stored in the memory 1005, and further perform the following operations:
[0061] The step of generating a visualization interface of the federated learning workflow includes:
[0062] The visualization interface is generated accordingly according to the operation process of each operation component and the display parameters.
[0063] Further, the processor 1001 may call a display program of the federated learning workflow interface stored in the memory 1005, and further perform the following operations:
[0064] Read the workflow configuration information carried by the preset federated learning task;
[0065] According to the workflow configuration information, configure an independently running work component in the federated learning workflow and specify a unilateral execution role of the independently running work component, and / or configure a jointly running work component in the federated learning workflow and specify a multi-party execution role of the jointly running work component;
[0066] According to the workflow configuration information, the operating parameter information of the independently running working component and / or the jointly running working component is configured in the federated learning workflow.
[0067] Further, the processor 1001 may call a display program of the federated learning workflow interface stored in the memory 1005, and after executing the output display of the visualization interface after adjusting the output display state to the output object, further perform the following operations:
[0068] A request instruction for viewing component output results is obtained, and the output results of the running component specified by the request instruction are displayed to the target output object.
[0069] Further, the processor 1001 may call a display program of the federated learning workflow interface stored in the memory 1005, and further perform the following operations:
[0070] Obtaining the operator of the operating component specified by the request instruction, and detecting whether the triggering end of the request instruction belongs to the operator;
[0071] If so, the trigger end is used as the target output object to output and display the output result.
[0072] Based on the above structure, various embodiments of the method for displaying the federated learning workflow interface of the present invention are proposed.
[0073] Please refer to Figure 2 , Figure 2 Schematic diagram of the flow chart of the first embodiment of the method for displaying a federated learning workflow interface of the present invention.
[0074] An embodiment of the present invention provides an embodiment of a method for displaying a federated learning workflow interface. It should be noted that although a logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in an order different from that shown here.
[0075] The method for displaying a federated learning workflow interface in an embodiment of the present invention is applied to the above-mentioned terminal device. The terminal device in the embodiment of the present invention can be a PC, a portable computer or other terminal device, and no specific limitation is made here.
[0076] The method for displaying the federated learning workflow interface in this embodiment includes:
[0077] Step S100, when the federated learning workflow is initiated, generating a visualization interface of the federated learning workflow.
[0078] In the federated machine learning system constructed by the coordinator, the demander and the data providers, when it is detected that the demander initiates the current federated machine learning workflow based on its own federated learning task processing needs, a visual interface of the current federated machine learning workflow is generated according to the federated learning workflow pre-configured by the current demander to process its own federated learning task.
[0079] It should be noted that, in this embodiment, based on the setting of the roles of the federated machine learning participants, the federated machine learning system contains a coordinator (arbiter), a demander (guest) and various data providers (host), and in the federated machine learning system, only the workflow of federated machine learning initiated by the demander guest is supported, that is, the demander guest completes the configuration of the entire federated learning workflow in advance and starts running it.
[0080] Step S200: detecting an output object of the visualization interface, and adjusting an output display state of the visualization interface according to the output object.
[0081] Further, step S200 includes:
[0082] Step S201, detecting the role type of the display output object in federated learning;
[0083] Step S202: adjusting the output display state of the visualization interface according to the role type.
[0084] After the demand side of each participant in the federated learning system initiates the current federated machine learning workflow and generates a corresponding visualization interface of the current federated machine learning workflow, the output objects to be output and displayed by the visualization interface in the current federated learning system are detected, so as to adjust the output display status of the visualization interface of the federated machine learning workflow accordingly according to the different role types of each output object in the current federated learning participant roles.
[0085] Further, step S202 includes:
[0086] Step S2021, filtering target working components in which the output object in the visualization interface is not visible according to the role type.
[0087] Specifically, for example, when it is necessary to output and display the visualization interface of the currently generated federated machine learning workflow initiated by the demander in the federated learning system to data provider a among multiple data providers in the current federated learning system, the target working components in which the current data provider a does not participate in the operation are extracted from the visualization interface that maps the entire operation process of all working components of the current federated machine learning. Similarly, when it is necessary to output and display the visualization interface to data provider b among multiple data providers, among all the working components displayed in the visualization interface, the target working components in which the data provider b does not participate in the operation are screened and extracted.
[0088] It should be noted that, in another embodiment, the running components in which data provider a (or data provider b) participates in running can also be used as target working components among all the running components output and displayed by the visualization interface. Therefore, when filtering the target working components in the visualization interface, the target working components in which data provider a (or data provider b) participates in running can be correspondingly filtered out from all the working components displayed in the visualization interface.
[0089] It should be noted that in this embodiment, based on the independent decision of each participant in the current federated learning system whether to view the visual interface demand of the federated machine learning workflow, the visual interface of the federated machine learning workflow initiated by the currently generated demander guest will be output and displayed accordingly. Before the output display, the output display status of the visual interface of the federated machine learning workflow will be adjusted accordingly according to the different role types of each output object in the current federated learning participant roles.
[0090] Step S2022: In the visualization interface, change the display parameters of the target working component to adjust the output display state of the visualization interface.
[0091] Specifically, for example, in Figure 3A In the visualization interface display scenario shown, in the visualization interface of the federated learning workflow that needs to be output and displayed to data provider a, except for the feature normalization component "federated_feature_scale" in which data provider a participates alone in running, and the longitudinal federated logistic regression component "hetero_federated_lr" in which data provider a participates in running together with other data providers (such as data provider b), the display status of other target running components in the current federated learning workflow visualization interface is modified and adjusted, that is, the display parameters such as the output display color of each target running component are adjusted and changed (grayed out or the display transparency is modified), and the name of each target running component is represented by "******"; further, in the example Figure 3BIn the visualization interface display scenario shown, in the visualization interface of the federated learning workflow that needs to be output and displayed to data provider b, except for the longitudinal federated feature binning component "hetero_federated_feature_binning" in which data provider b jointly with other data providers (other data providers except data provider a among the multiple data providers in the current federated learning system), and the longitudinal federated logistic regression component "hetero_federated_lr" in which data provider b jointly with other data providers (such as data provider a) participates in running, the display status of other target running components in the current federated learning workflow visualization interface is modified and adjusted, that is, the display parameters such as the output display color of each target running component are adjusted and changed (grayed out or the display transparency is modified), and the name of each target running component is represented by "******".
[0092] Furthermore, in another embodiment, when the visualization interface output of the current federated machine learning workflow is displayed to the demand-side guest in the current federated learning system, all running components in the current visualization interface can be output to the demand-side guest using display parameters such as different output display colors.
[0093] like Figure 3C In the visualization interface display scenario shown, when the visualization interface output of the current federated machine learning workflow is displayed to the demand-side guest (marked as: demander g) in the current federated learning system, since the demand-side guest has pre-configured and initiated the current federated machine learning workflow, the demander can view all the working components output and displayed in the visualization interface of the current federated machine learning workflow, that is, all the running components in the current visualization interface only need to adjust the display parameters such as the output display color of the working components that are not run by the demand-side guest, without the need to represent the names of the working components that are not run by the demand-side guest with "******".
[0094] It should be noted that, in this embodiment, the demand side guest in the federated learning system can see the entire structure of the federated learning workflow. However, for the components in the federated learning workflow that are not run by the demand side guest, such as the "feature normalization" component - federated_feature_scale_ (run separately by a data provider in the federated learning system), the component not run by the demand side guest will be marked in gray and displayed in output; the data provider host in the federated learning system cannot view the component information of the entire federated learning workflow, but can only view the components run only by itself, and use clear component status marks for output display, while the remaining components not run by itself are all represented by "******", and the number of input and output ports of other components not run by itself cannot be viewed, so as to avoid the data provider guessing the running component based on the number of ports unique to the running component; the coordinator arbiter in the federated learning system can only view the status of the working components jointly run by other participants (such as guests). For example, in Figure 3A and Figure 3B In the visualization interface output display scenarios shown separately, data provider a can only see the "feature normalization" component - federated_feature_scale, and the "longitudinal federated logistic regression" component - hetero__federated_lr, in which it participates in running; while data provider b can only see the "longitudinal federated feature binning" component - hetero_federated_feature_binning and the "longitudinal federated logistic regression" component - hetero__federated_lr, in which it participates in running.
[0095] It should be noted that, in this embodiment, the types of display parameters of each target running component that need to be adjusted and changed include but are not limited to display color, shape, etc. It should be understood that the display method of the federated learning workload interface of the present invention does not limit the type of change in the display parameters of the running component and the degree of change.
[0096] Step S300, outputting and displaying the visual interface after adjusting the output display state to the output object.
[0097] After completing the adjustment of the display status of the visualization interface of the federated machine learning process initiated by the demander, the visualization interface output adjusted according to the role type of the current participant in the federated learning system will be displayed to the current participant.
[0098] Specifically, for example, in the interface display application scenario shown in Figure 3, the display parameters of each target running component in the visualization interface will be adjusted and changed based on the operation status of each component in the current federated machine learning process in which data provider a participates (i.e., the running components participating in the operation alone or jointly), and the visualization interface output after adjustment by data provider a will be displayed to the terminal device where data provider a is located.
[0099] In this embodiment, in the federated machine learning system constructed by the coordinator, the demander and each data provider, when it is detected that the demander starts to initiate the current federated machine learning workflow based on its own federated learning task processing needs, a visualization interface of the current federated machine learning workflow is generated according to the federated learning workflow required for processing its own federated learning task that is pre-configured by the current demander. After the demander of each participant in the federated learning system starts to initiate the current federated machine learning workflow and generates a visualization interface of the current federated machine learning workflow accordingly, the output objects to be output and displayed by the visualization interface in the current federated learning system are detected, so as to adjust the output display state of the visualization interface of the federated machine learning workflow according to the different role types of each output object in the current federated learning participant role. After completing the adjustment of the display state of the visualization interface of the federated machine learning process initiated by the demander, the visualization interface adjusted according to the role type of the current participant in the federated learning system is output and displayed to the current participant.
[0100] It is achieved that when displaying a federated learning workflow involving multiple parties, different output display states of the visualization interface are displayed for different participants, ensuring that the machine learning process of each participant will not be known to other participants, thereby achieving the purpose of protecting the data privacy and data security of each federated learning participant in the process of displaying the federated learning workflow.
[0101] Furthermore, based on the first embodiment of the method for displaying a federated learning workflow interface described above, a second embodiment of the method for displaying a federated learning workflow interface of the present invention is proposed.
[0102] In a second embodiment of the method for displaying a federated learning workflow interface of the present invention, in step S100 of the first embodiment above, when the federated learning workflow is initiated, before generating a visualization interface of the federated learning workflow, the method for displaying a federated learning workflow interface of the present invention further includes:
[0103] Step S400: configuring each operating component in the federated learning workflow based on a preset federated learning task.
[0104] Obtain the preset federated learning task based on which the demander initiates the visual interface of the current federated machine learning workflow, read the federated learning workflow configuration information of the preset federated learning task, and configure each running component in the current federated learning workflow based on the workflow configuration information.
[0105] Furthermore, in step S100, the step of generating a visualization interface of the federated learning workflow includes:
[0106] Step S101, generating the visualization interface accordingly according to the operation process of each operation component and the display parameters.
[0107] When the demand side initiates the current federated machine learning workflow based on its own federated learning task processing needs, a visual interface mapping the current federated machine learning workflow is generated based on the input and output processes of each running component required to process its own federated learning task pre-configured by the current demand side, as well as the display parameters of each running component (i.e., output display color, display content, etc.).
[0108] Furthermore, step S400 includes:
[0109] Step S401, reading the workflow configuration information carried by the preset federated learning task.
[0110] Step S402: According to the workflow configuration information, an independently running work component is configured in the federated learning workflow, and a unilateral execution role of the independently running work component is specified, and / or, a jointly running work component is configured in the federated learning workflow, and a multi-party execution role of the jointly running work component is specified.
[0111] Specifically, for example, when configuring a federated machine learning workflow, the demander guest can, based on the data processing task, specify a data provider host among multiple data providers in the current federated learning system to independently run the feature operation components required for the execution of the current data processing task. Alternatively, on the other hand, the demander guest can, based on the model training processing task, specify a data provider host among multiple data providers and an arbiter to jointly run the feature operation components required for the execution of the current model training processing task. Alternatively, the demander guest can, based on the data processing task, specify a data provider host among multiple data providers in the current federated learning system to independently run the feature operation components required for the execution of the current data processing task, and can also specify a data provider host among multiple data providers and an arbiter to jointly run the feature operation components required for the execution of the current model training processing task.
[0112] Furthermore, in another embodiment, the demander guest may also specify a data provider host among multiple data providers in the current federated learning system to independently run the current component (i.e., the "feature normalization" component) based on the feature work component (e.g., the "feature normalization" component - federated_feature_scale); or, the demander guest may also specify a data provider host among multiple data providers and the coordinator arbiter to jointly run the current feature work component based on the feature work component (e.g., the "longitudinal federated logistic regression" component - hetero_federated_lr).
[0113] Step S403: configuring the operating parameter information of the independently running working component and / or the jointly running working component in the federated learning workflow according to the workflow configuration information.
[0114] After the demand-side guest specifies the various feature operation components that are independently and / or jointly executed by the executor based on the federated learning workflow task configuration information of the federated learning processing task, the operation parameter information of each feature operation component negotiated and agreed upon with the coordinator arbiter and the data provider hosts in the current federated learning system will be configured on the corresponding feature operation components, thereby completing the configuration of the current federated learning workflow.
[0115] In this embodiment, the federated learning processing task based on which the demander starts the visual interface for initiating the current federated machine learning workflow is obtained, and the task type in the task attributes of the federated learning processing task is read. The demander configures the federated learning workflow required to process the federated learning task based on the task type and characteristic working components of the read federated learning processing task, that is, specifies that the participants in the federated learning system run the current federated learning workflow independently or jointly.
[0116] Furthermore, based on the first embodiment of the method for displaying a federated learning workflow interface described above, a third embodiment of the method for displaying a federated learning workflow interface of the present invention is proposed.
[0117] In a third embodiment of the method for displaying a federated learning workflow interface of the present invention, in step S300 of the first embodiment above, after the output of the visual interface after adjusting the output display state is displayed to the output object, the method for displaying a federated learning workflow interface of the present invention further includes:
[0118] Step S500: obtaining a request instruction for viewing component output results, and displaying the output results of the running component specified by the request instruction to a target output object.
[0119] Detect and obtain any participant in the current federated learning system, trigger the generation of a request instruction for viewing the output results of any running component in the federated machine learning workflow, detect and determine all the running parties in the current federated learning system that have the authority to view the output results of the running component specified by the request instruction, and when it is detected that the participant who triggered the generation of the request instruction is a member of the running parties, output and display the participant as the target output object of the output result of the current running component.
[0120] Furthermore, step S500 includes:
[0121] Step S501, obtaining the operator of the operating component specified by the request instruction, and detecting whether the triggering end of the request instruction belongs to the operator.
[0122] Detect all participants in the current federated learning system (i.e., the coordinator, each demander, and all data providers) who have the authority to view the output results of the specified running component in response to the obtained request instruction for viewing the output results of the running component, and further detect whether the participant who currently triggers the generation of the request instruction (which can be the coordinator, the demander, or any data provider) is one of the detected operators.
[0123] Step S502: taking the trigger end as the target output object to output and display the output result.
[0124] When it is detected that the participant who currently triggers the generation request instruction belongs to one of the detected operating parties, the participant is used as the target output object of the output result of the current operating component, and the output result is output and displayed to the participant who currently triggers the generation request instruction.
[0125] Furthermore, in another embodiment, when it is detected that the participant who currently triggers the generation request instruction does not belong to one of the detected operating parties, the participant is abandoned as the target output object of the output result of the current operating component, and the corresponding prompt information is output to the participant who currently triggers the generation request instruction (for example, "does not have viewing permission, cannot know the component operation status", etc.).
[0126] It should be noted that, in this embodiment, the output results of each running component displayed in the visual output of the federated machine learning workflow are displayed differently according to the different roles of the participants in the current federated learning system (i.e., the coordinator arbiter, the demander guest, and the data provider host). That is, the output results (model output, data output, etc.) contained in the output results of each running component are different when output by different roles in the federated learning system. That is, the output results of each running component displayed to each participant in the visual output of the federated machine learning are different in terms of viewing permissions and viewing content, thereby further protecting the data privacy and data security of each participant in the federated learning.
[0127] Specifically, for example, for a component that runs independently on one side and is displayed in a federated machine learning visualization output, the output result of the component is output only to the sole executor of the component; while for a component that runs jointly, the output result of the component can be output and displayed to two or more parties that are jointly executed, that is, part of the output result of the component that each executor participates in the operation can be output and displayed to each executor (for example, the "vertical federated logistic regression" component - hetero__federated_lr can output and display part of the model output of the local longitudinal federated model to each executor, without outputting part of the model output of the local logistic regression model of other participants in the joint operation, or the output result of the component that part of the data provider host participates in the operation can be output and displayed to the demand side guest in the federated learning system (for example, the "vertical federated feature binning" component - hetero_federated_feature_binnin g. For the demand side guest in the federated learning system, in addition to being able to output and display the binning details obtained based on the characteristics of the demand side guest (part of the output results of the vertical federated feature binning component), it is also possible to output and display the other binning details except the feature binning split points based on the characteristics of each data provider host participating in the current "vertical federated feature binning" component - hetero_federated_feature_binning operation (part of the output results of the vertical federated feature binning component). For each data provider host in the federated learning system, only the feature binning split points obtained by the current "vertical federated feature binning" component - hetero_federated_feature_binning operation based on the characteristics of each data provider host can be output and displayed, but the feature binning details, such as iv value (information value, amount of information or information value), woe value (weight of evidence), etc., so as to avoid the data provider host from using the feature binning details to infer the key information and output results of the "vertical federated feature binning" component - hetero_federated_feature_binning run by the demander guest in the current federated learning system.
[0128] In this embodiment, the output results of each running component displayed in the visualization output of the federated machine learning workflow are displayed differently according to the different roles of the participants in the current federated learning system (i.e., the coordinator arbiter, the demander guest, and the data provider host), so that the output results (model output, data output, etc.) contained in the output results of each running component are different when output and displayed by different roles in the federated learning system. That is, the output results of each running component displayed to each participant in the visualization output of the federated machine learning are different in viewing permissions and viewing content, thereby further protecting the data privacy and data security of each participant in the federated learning.
[0129] In addition, please refer to Figure 4 The embodiment of the present invention further provides a display device for a federated learning workflow interface. The display device for a federated learning workflow interface of the present invention includes:
[0130] A generation module, used to generate a visual interface of the federated learning workflow when the federated learning workflow is initiated;
[0131] An adjustment module, used for detecting an output object of the visualization interface, and adjusting an output display state of the visualization interface according to the output object;
[0132] The display module is used to display the output of the visual interface after adjusting the output display state to the output object.
[0133] Preferably, the adjustment module includes:
[0134] A detection unit, used to detect the role type of the display output object in federated learning;
[0135] An adjustment unit is used to adjust the output display state of the visualization interface according to the role type.
[0136] Preferably, the adjustment unit further includes:
[0137] A screening subunit, used for screening target working components that are not visible to the output object in the visualization interface according to the role type;
[0138] The changing subunit is used to change the display parameters of the target working component in the visualization interface to adjust the output display state of the visualization interface.
[0139] Preferably, the display device of the federated learning workflow interface of the present invention further includes:
[0140] A configuration module, used to configure each running component in the federated learning workflow based on a preset federated learning task,
[0141] The generating module comprises:
[0142] The generating unit is used to generate the visualization interface according to the operation process of each operation component and the display parameters.
[0143] Preferably, the configuration module includes:
[0144] A reading unit, used to read the workflow configuration information carried by the preset federated learning task;
[0145] A first configuration unit is used to configure an independently running work component in the federated learning workflow according to the workflow configuration information, and specify a unilateral execution role of the independently running work component, and / or configure a jointly running work component in the federated learning workflow, and specify a multi-party execution role of the jointly running work component;
[0146] The second configuration unit is used to configure the operating parameter information of the independently running working component and / or the jointly running working component in the federated learning workflow according to the workflow configuration information.
[0147] Preferably, the display module of the display device of the federated learning workflow interface of the present invention is also used to obtain a request instruction for viewing the output result of the component, and display the output result of the running component specified by the request instruction to the target output object.
[0148] Preferably, the display module comprises:
[0149] A detection and confirmation unit, used to obtain the operator of the operation component specified by the request instruction, and detect whether the triggering end of the request instruction belongs to the operator;
[0150] An output display unit is used to use the trigger end as the target output object to output and display the output result.
[0151] In addition, an embodiment of the present invention further proposes a computer-readable storage medium, which is applied to a computer. The computer-readable storage medium may be a non-volatile computer-readable storage medium, and a display program for a federated learning workflow interface is stored on the computer-readable storage medium. When the display program for the federated learning workflow interface is executed by a processor, the steps of the method for displaying the federated learning workflow interface as described above are implemented.
[0152] Among them, the steps implemented when the display program of the federated learning workflow interface running on the processor is executed can refer to the various embodiments of the display method of the federated learning workflow interface of the present invention, and will not be repeated here.
[0153] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or system including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or system. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or system including the element.
[0154] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.
[0155] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, a magnetic disk, or an optical disk), and includes a number of instructions for enabling a terminal device (which can be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to execute the methods described in each embodiment of the present invention.
[0156] The above are only preferred embodiments of the present invention, and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.
Claims
1. A method for displaying a federated learning workflow interface, characterized in that: The method for displaying the federated learning workflow interface includes: When a federated learning workflow is initiated, generating a visualization interface of the federated learning workflow; Detecting an output object of the visualization interface, and adjusting an output display state of the visualization interface according to the output object; Outputting the visual interface after adjusting the output display state to the output object; The step of adjusting the output display state of the visualization interface according to the output object comprises: Detecting a role type of the output object in federated learning; According to the role type, filter the target working components in the visualization interface that are invisible to the output object; In the visualization interface, display parameters of the target working component are changed to adjust the output display state of the visualization interface.
2. The method for displaying a federated learning workflow interface according to claim 1, characterized in that: Before the step of generating a visualization interface of the federated learning workflow when the federated learning workflow is initiated, the step further includes: Configure each running component in the federated learning workflow based on a preset federated learning task; The step of generating a visualization interface of the federated learning workflow includes: The visualization interface is generated accordingly according to the operation process of each operation component and the display parameters.
3. The method for displaying a federated learning workflow interface according to claim 2, wherein: The step of configuring each running component in the federated learning workflow based on the preset federated learning task includes: Read the workflow configuration information carried by the preset federated learning task; According to the workflow configuration information, configure an independently running work component in the federated learning workflow and specify a unilateral execution role of the independently running work component, and / or configure a jointly running work component in the federated learning workflow and specify a multi-party execution role of the jointly running work component; According to the workflow configuration information, the operating parameter information of the independently running working component and / or the jointly running working component is configured in the federated learning workflow.
4. The method for displaying a federated learning workflow interface according to claim 1, wherein: After the step of outputting and displaying the visual interface after adjusting the output display state to the output object, the method further includes: A request instruction for viewing component output results is obtained, and the output results of the running component specified by the request instruction are displayed to the target output object.
5. The method for displaying a federated learning workflow interface according to claim 4, characterized in that: The step of displaying the output result of the running component specified by the request instruction to the target output object includes: Obtaining the operator of the operating component specified by the request instruction, and detecting whether the triggering end of the request instruction belongs to the operator; If so, the trigger end is used as the target output object to output and display the output result.
6. A display device for a federated learning workflow interface, characterized in that: The display device of the federated learning workflow interface includes: A generation module, used to generate a visual interface of the federated learning workflow when the federated learning workflow is initiated; An adjustment module, used for detecting an output object of the visualization interface, and adjusting an output display state of the visualization interface according to the output object; A display module, used to display the output of the visual interface after adjusting the output display state to the output object; The adjustment module comprises: A detection unit, used to detect the role type of the output object in federated learning; An adjustment unit, configured to adjust an output display state of the visualization interface according to the role type; The adjustment unit comprises: A screening subunit, used for screening target working components that are not visible to the output object in the visualization interface according to the role type; The changing subunit is used to change the display parameters of the target working component in the visualization interface to adjust the output display state of the visualization interface.
7. A terminal device, characterized in that: The terminal device includes: a memory, a processor, and a display program for a federated learning workflow interface stored in the memory and executable on the processor. When the display program for the federated learning workflow interface is executed by the processor, the steps of the method for displaying a federated learning workflow interface as described in any one of claims 1 to 5 are implemented.
8. A storage medium, characterized in that: The storage medium stores a computer program, which, when executed by a processor, implements the steps of the method for displaying a federated learning workflow interface according to any one of claims 1 to 5.
