Trust model evaluation method and device

Abstract

The invention discloses a trust model evaluation method and device, and relates to the field of information security. The method comprises the steps that trust evaluation is conducted on a first entity based on a trust model, trust evaluation results of the first entity are obtained, the trust evaluation results comprise at least one group, each group of trust evaluation results corresponds to one first index, and the first indexes are used for indicating evaluation dimensions of the trust model; and according to the trust evaluation result, evaluating the trust model to obtain an evaluation result of the trust model, the evaluation result of the trust model being used for indicating the trust evaluation capability of the trust model in the dimension of the first index. According to the method, the trust evaluation capability of the trust model can be evaluated according to the dimension of the first index for the entity in a specific application scene, such as the first entity, and a decision basis is provided for selection of the trust model.

Description

Technical Field

[0001] This application relates to the field of information security, and in particular, to a method and device for evaluating a trust model. Background Art

[0002] Trust assessment generally refers to evaluating the trust level of entities in an application scenario. By performing trust assessment on entities, it can be understood whether an entity is trustworthy, whether an entity can interact, or whether the data of an entity is available, etc. For example, in some communication-related application scenarios, an entity can refer to an interaction object or a communication node in a communication network, such as a network device or a terminal device. According to the trust assessment result of a certain communication node in the communication network, it can be determined whether the communication node is secure, such as whether the data of the communication node is trustworthy, to detect the security and reliability of the communication network.

[0003] The methods of trust assessment can include: using a trust assessment model to perform trust assessment on an entity and outputting the trust assessment result of the entity. Among them, the trust assessment model is also called a trust model, and the trust assessment result can refer to the trust level of the entity. Different trust models may have different granularities in outputting trust assessment results. For example, the trust assessment results output by some trust models may only include trustworthy or untrustworthy, with a larger (or coarser) granularity, and there are also some trust models whose trust assessment results may be divided according to multiple levels of trust, with a smaller (or finer) granularity.

[0004] For entities in a specific application scenario, currently, generally according to manual experience, a trust model is artificially selected to perform trust assessment on the entity. However, the trust assessment results output by the artificially selected trust model may have too large a granularity, resulting in low accuracy of the trust assessment results, or may have too small a granularity, resulting in increased unnecessary computing costs. Summary of the Invention

[0005] This application provides a method and device for evaluating a trust model, which can evaluate the trust assessment ability of a trust model according to the dimension of a first indicator for an entity, such as a first entity, in a specific application scenario, and provide a decision basis for the selection of the trust model.

[0006] In a first aspect, this application provides a method for evaluating a trust model. The method includes: performing trust assessment on a first entity based on a trust model to obtain a trust assessment result of the first entity. The trust assessment result includes at least one group, and each group of trust assessment results corresponds to a first indicator. The first indicator is used to indicate the evaluation dimension of the trust model; according to the trust assessment result, evaluating the trust model to obtain an evaluation result of the trust model. The evaluation result of the trust model is used to indicate the trust assessment ability of the trust model within the dimension of the first indicator.

[0007] Exemplarily, the method is applied to an electronic device. For example, the method is executed by the electronic device, or by a device (such as a chip or a functional module) built into the electronic device.

[0008] This trust model evaluation method can evaluate the trust evaluation ability of the trust model for an entity, such as a first entity, in a specific application scenario according to the dimension of a first metric. By evaluating the trust evaluation ability of the trust model, it can assist in selecting a more suitable trust model for the trust evaluation of entities in the application scenario, or assist the user in judging whether the selected trust model is appropriate. In other words, it can provide a decision basis for the selection of the trust model. For example, the user can judge whether the trust evaluation granularity is appropriate based on the evaluation result of the trust model, or select a trust model with a more appropriate trust evaluation granularity.

[0009] In a possible design, the trust evaluation of the first entity based on the trust model to obtain the trust evaluation result of the first entity includes: obtaining the first metric, the scenario information of the application scenario, and the entity information of the first entity; generating an evaluation environment according to the first metric, the scenario information of the application scenario, and the entity information of the first entity, where the evaluation environment indicates the environment in which the first entity runs in the application scenario; invoking the trust model, and performing trust evaluation on the first entity based on the evaluation environment to obtain the trust evaluation result of the trust model for the first entity.

[0010] In this design, for a specific application scenario and the first entity in the application scenario, after generating the evaluation environment, further evaluating the trust evaluation ability of the trust model based on the evaluation environment can enable the evaluation environment to be flexibly adjusted according to the application scenario and the first entity. This method can be more flexibly applicable to the evaluation of trust models in more application scenarios. In other words, the implementation framework of this method can be self-flexibly adjusted according to the changes in the application scenario and the first entity, and has strong adaptability.

[0011] In a possible design, the obtaining of the first metric, the scenario information of the application scenario, and the entity information of the first entity includes: obtaining the scenario information of the application scenario and the entity information of the first entity; and obtaining the first metric based on the scenario information of the application scenario and the entity information of the first entity.

[0012] Exemplarily, the ways to obtain the scenario information of the application scenario and the entity information of the first entity may include: receiving the scenario information of the application scenario and the entity information of the first entity input by the user, or obtaining the scenario information of the application scenario and the entity information of the first entity from a specified data interface.

[0013] In some implementation manners, the first metric used to generate the evaluation environment may be predefined, preconfigured, or configured. For example, evaluation metrics matching the application scenario and the entity may be configured to obtain the mapping relationship between the evaluation metrics, the application scenario, and the entity. Based on the scenario information of the application scenario and the entity information of the first entity, obtaining the first metric may include: determining the evaluation metric matching the application scenario and the first entity according to this mapping relationship, such as the so-called target evaluation metric, and selecting at least one metric from the target evaluation metrics as the first metric.

[0014] In this implementation manner, the first metric may be related to the application scenario and / or the first entity, and it can be customized to select the first metric for the current application environment and the first entity according to the user's design to evaluate the trust model, which can improve the accuracy of the trust model evaluation result and enable the trust model evaluation result to reflect the performance of the trust model from a more accurate and effective dimension.

[0015] In some other implementation manners, the first metric used to generate the evaluation environment may also be dynamically determined according to the scenario information and the entity information of the first entity. For example, based on the scenario information of the application scenario and the entity information of the first entity, obtaining the first metric may include: through deep learning or machine learning, the pre-trained neural network model dynamically determines the evaluation metric matching the application scenario and the first entity according to the scenario information and the entity information of the first entity, such as the so-called target evaluation metric, and selecting at least one metric from the target evaluation metrics as the first metric.

[0016] This design can be customized to dynamically select the first metric for the current application environment and the first entity in a targeted manner to evaluate the trust model, which can improve the accuracy of the trust model evaluation result, enhance the adaptability of the trust model evaluation to different application scenarios, increase the flexibility of the evaluation, and enable the trust model evaluation result to reflect the performance of the trust model from a more accurate and effective dimension.

[0017] In yet another possible design, the first metric used to generate the evaluation environment may also be one or more first metrics selected by the user from the obtained first metrics based on the scenario information of the application scenario and the entity information of the first entity. For example, the user may perform an operation to select the first metric, and the method further includes: in response to the user's operation of selecting the first metric, using the first metric indicated by the selection operation as the first metric used to generate the evaluation environment.

[0018] In this design, the first metric obtained based on the scenario information of the application scenario and the entity information of the first entity may be referred to as the alternative first metric, and the first metric selected by the user from the alternative first metrics may be referred to as the target first metric.

[0019] This design can customize the selection of the first metric for the current application environment and the first entity, and at the same time, it also allows users to adjust the first metric according to their specific needs. For example, users can select more meaningful or relevant first metrics according to their needs, which can further improve the value of the evaluation results of the trust model for users. In addition, by screening and selecting more relevant first metrics, the noise brought by irrelevant or secondary metrics can also be reduced, further improving the accuracy of the evaluation results.

[0020] Alternatively, users can also select some specific metrics for evaluation according to the need to save evaluation time and resources, so as to improve the evaluation efficiency.

[0021] In a possible design, obtaining the first metric based on the scenario information of the application scenario and the entity information of the first entity includes: fusing the scenario information of the application scenario and the entity information of the first entity to obtain a fused feature; and obtaining the first metric based on the fused feature.

[0022] For example, the scenario features of the scenario information of the application scenario and the entity features of the entity information of the first entity can be extracted, and the scenario features and the entity features are fused to obtain a fused feature.

[0023] In this design, obtaining the first metric based on the fused feature can further improve the effectiveness or accuracy of the first metric, and select a more suitable first metric for the application scenario and the first entity for the evaluation of the trust model.

[0024] In a possible design, generating an evaluation environment according to the first metric, the scenario information of the application scenario, and the entity information of the first entity in the application scenario includes: successively for each of the first metrics, generating an evaluation environment corresponding to the first metric according to the first metric, the scenario information of the application scenario, and the entity information of the first entity in the application scenario.

[0025] Correspondingly, calling the trust model to perform a trust evaluation on the first entity based on the evaluation environment and outputting the trust evaluation result of the first entity includes: performing a trust evaluation on the first entity based on the evaluation environment corresponding to the first metric and outputting the trust evaluation result of the first entity.

[0026] That is, in this design, each first indicator corresponds to an evaluation environment, and the evaluation environments corresponding to different first indicators are different. This design can implement the process of generating an evaluation environment according to the first indicator and performing an evaluation based on the evaluation environment in a serial manner for multiple first indicators. For example, for one first indicator, the trust evaluation ability of the trust model can be evaluated first; after completing the evaluation of the trust model corresponding to one first indicator, the evaluation of the trust model corresponding to the next first indicator can be continued until the evaluation of the trust model corresponding to all first indicators is completed.

[0027] In this design, for multiple first indicators, implementing the process of generating an evaluation environment according to the first indicator and performing an evaluation based on the evaluation environment in a serial manner, focusing on a single indicator (evaluating only one specific first indicator at a time) can make the evaluation process more concise, clear, and easy to understand and interpret. In addition, for multiple first indicators, implementing the process of generating an evaluation environment according to the first indicator and performing an evaluation based on the evaluation environment in a serial manner can also enhance the evaluation depth and improve the evaluation accuracy. For example, resources and attention can be concentrated on one indicator, avoiding the noise and interference that may be brought by multiple indicators, and more clearly understanding and analyzing the importance and role of each indicator.

[0028] In a possible design, the generated evaluation environment can include an initial evaluation environment and at least one other evaluation environment; the initial evaluation environment is generated according to the first indicator, scenario information, and the entity information of the first entity; the other evaluation environment can be obtained by adjusting the parameters in the initial evaluation environment.

[0029] For example, in this design, the method further includes: adjusting the parameters of the evaluation environment to obtain at least one updated evaluation environment; calling the trust model, and performing a trust evaluation on the first entity based on the updated evaluation environment, and outputting the trust evaluation result of the first entity. Among them, the evaluation environment is the initial evaluation environment, and the updated evaluation environment is the other evaluation environment.

[0030] In this design, the evaluation environment can be extended to provide a richer evaluation environment for the evaluation of the trust model and improve the accuracy of the trust model evaluation. For example, the evaluation of the trust model can be performed based on multiple evaluation environments such as the initial evaluation environment and other evaluation environments, and the performance of the trust model can be comprehensively evaluated according to the evaluation results obtained from different evaluation environments.

[0031] In a possible design, the application scenarios include at least one of the following: network security scenarios, intelligent system scenarios, online platform scenarios, complex decision-making environment scenarios, Internet of Things (IoT) scenarios, cloud computing scenarios, edge computing scenarios, network slicing and virtualized network scenarios, software-defined network scenarios, blockchain scenarios, communication network scenarios; the first entity includes at least one of the following: hardware devices or modules, applications or software modules, and data of hardware and / or software interactions.

[0032] In a possible design, the evaluation result of the trust model is also used to indicate the update strategy of the trust model.

[0033] In this design, the evaluation result of the trust model can indicate the update strategy of the trust model, which can point out the improvement direction (dimension) of the trust model for developers or users, or provide more detailed model optimization suggestions for users, enabling users to further optimize or adjust the trust model according to the evaluation result of the trust model to improve the performance of the trust model.

[0034] In a possible design, the method further includes: updating the trust model according to the evaluation result of the trust model.

[0035] In this design, updating the trust model according to the evaluation result can improve the trust evaluation ability of the trust model.

[0036] In a possible design, the evaluation result of the trust model is a semantic trust evaluation report.

[0037] Exemplarily, the trust evaluation result corresponding to the first metric can be input into a large language model (LLM), and the LLM parses the trust evaluation result corresponding to the first metric and outputs a semantic trust evaluation report as the evaluation result of the trust model.

[0038] In this design, using the semantic trust evaluation report as the evaluation result of the trust model enables users to better understand the trust evaluation ability of the trust model within the dimension of the first metric, and more intuitively understand from which dimensions the trust model can be improved, and / or in what ways the trust model can be improved.

[0039] In a possible design, the first metric includes one or more of the following metrics: data security, comprehensiveness of trust evaluation results, usability of the trust model, functionality of the trust model, robustness of the trust model, neutrality of trust evaluation results, and interpretability of trust evaluation results.

[0040] This application does not limit the specific type of the first metric.

[0041] In a second aspect, the present application provides a trust model evaluation device, which has the function of implementing the method described in the first aspect above. The function can be implemented by hardware or by hardware executing corresponding software. The hardware or software includes one or more units or modules corresponding to the functions of the method described in the first aspect above. For example, a trust evaluation unit, a model evaluation unit, etc.

[0042] Among them, the trust evaluation unit is used to perform a trust evaluation on the first entity based on the trust model to obtain a trust evaluation result of the first entity. The trust evaluation result includes at least one group, and each group of trust evaluation results corresponds to a first metric, and the first metric is used to indicate the evaluation dimension of the trust model.

[0043] The model evaluation unit is used to evaluate the trust model according to the trust evaluation result to obtain an evaluation result of the trust model. The evaluation result of the trust model is used to indicate the trust evaluation ability of the trust model within the dimension of the first metric.

[0044] In a possible design, the trust evaluation unit is specifically used to: obtain the first metric, the scenario information of the application scenario, and the entity information of the first entity; generate an evaluation environment according to the first metric, the scenario information of the application scenario, and the entity information of the first entity, where the evaluation environment indicates the environment in which the first entity runs in the application scenario; call the trust model and perform a trust evaluation on the first entity based on the evaluation environment to obtain the trust evaluation result of the trust model for the first entity.

[0045] In a possible design, the trust evaluation unit is specifically used to: obtain the scenario information of the application scenario and the entity information of the first entity; obtain the first metric based on the scenario information of the application scenario and the entity information of the first entity.

[0046] In a possible design, the first metric used to generate the evaluation environment can be predefined, or pre-configured, or configured. For example, evaluation metrics can be configured for the application scenario and the entity to obtain a mapping relationship between the evaluation metrics, the application scenario, and the entity. The trust evaluation unit is specifically used to: determine the evaluation metrics that match the application scenario and the first entity according to this mapping relationship, such as called target evaluation metrics, and select at least one metric from the target evaluation metrics as the first metric.

[0047] In another possible design, the first metric used to generate the evaluation environment can also be dynamically determined according to the scenario information and the entity information of the first entity. For example, the trust evaluation unit is further used to dynamically determine, by means of deep learning or machine learning, the evaluation metrics that match the application scenario and the first entity, such as called target evaluation metrics, using a pre-trained neural network model according to the scenario information and the entity information of the first entity, and select at least one metric from the target evaluation metrics as the first metric.

[0048] In another possible design, the first metric used to generate the evaluation environment can also be one or more first metrics selected by the user from the obtained first metrics based on the scenario information of the application scenario and the entity information of the first entity. For example, the trust evaluation unit is further configured to, in response to the user's selection operation on the first metric, use the first metric indicated by the selection operation as the first metric used to generate the evaluation environment.

[0049] In one possible design, the trust evaluation unit is specifically configured to: fuse the scenario information of the application scenario and the entity information of the first entity to obtain a fused feature; and obtain the first metric based on the fused feature.

[0050] In one possible design, the trust evaluation unit is specifically configured to, for each of the first metrics in sequence, generate an evaluation environment corresponding to the first metric according to the first metric, the scenario information of the application scenario, and the entity information of the first entity in the application scenario; perform a trust evaluation on the first entity based on the evaluation environment corresponding to the first metric, and output the trust evaluation result of the first entity.

[0051] That is, in this design, each first metric corresponds to an evaluation environment, and the evaluation environments corresponding to different first metrics are different.

[0052] In one possible design, the generated evaluation environment may include an initial evaluation environment and at least one other evaluation environment; the initial evaluation environment is generated according to the first metric, the scenario information, and the entity information of the first entity; the other evaluation environment may be obtained by adjusting the parameters in the initial evaluation environment. For example, the trust evaluation unit is further configured to adjust the parameters of the evaluation environment to obtain at least one updated evaluation environment; call the trust model, perform a trust evaluation on the first entity based on the updated evaluation environment, and output the trust evaluation result of the first entity. Here, the evaluation environment is the initial evaluation environment, and the updated evaluation environment is the other evaluation environment.

[0053] In one possible design, the application scenario includes at least one of the following: network security scenario, intelligent system scenario, online platform scenario, complex decision-making environment scenario, Internet of Things scenario, cloud computing scenario, edge computing scenario, network slicing and virtualized network scenario, software-defined network scenario, blockchain scenario, communication network scenario; the first entity includes at least one of the following: a hardware device or module, an application or software module, and data of hardware and / or software interaction.

[0054] In one possible design, the evaluation result of the trust model is also used to indicate the update strategy of the trust model.

[0055] In a possible design, the device further includes an updating unit configured to update the trust model according to the evaluation result of the trust model.

[0056] In a possible design, the evaluation result of the trust model is a semantic trust evaluation report.

[0057] In a possible design, the first metric includes one or more of the following metrics: security of data, comprehensiveness of the trust evaluation result, usability of the trust model, functionality of the trust model, robustness of the trust model, neutrality of the trust evaluation result, and interpretability of the trust evaluation result.

[0058] In a third aspect, the present application further provides a trust model evaluation device, including a processor configured to execute computer instructions stored in a memory. When the computer instructions are executed, the device is caused to execute the method described in the first aspect or any possible design of the first aspect. Alternatively, the processor is configured to execute the method described in the first aspect or any possible design of the first aspect.

[0059] In a fourth aspect, the present application further provides a trust model evaluation device, including a processor and an interface circuit. The processor is configured to communicate with other devices through the interface circuit and execute the method described in the first aspect or any possible design of the first aspect.

[0060] The trust model evaluation device described in the second to fourth aspects above may be an electronic device or a device (e.g., a chip) built into an electronic device.

[0061] In a fifth aspect, the present application further provides a computer-readable storage medium, including computer software instructions (or referred to as instructions). When the computer software instructions are run, the method described in the first aspect or any possible design of the first aspect is implemented. For example, when the computer software instructions run in an electronic device or a device (e.g., a chip) built into an electronic device, the electronic device is caused to execute the method described in the first aspect or any possible design of the first aspect.

[0062] It can be understood that the beneficial effects that can be achieved by the second to fifth aspects provided above may refer to the beneficial effects in the first aspect and any of its possible designs, which will not be elaborated here.

[0063] In a sixth aspect, the present application further provides a trust model evaluation device, including a transceiver unit and a processing unit. The transceiver unit may be configured to send and receive information or communicate with other network elements (such as a terminal device or a network device, or other electronic devices or devices). The processing unit may be configured to process data. The device may implement the method described in the first aspect and any of its possible designs through the transceiver unit and the processing unit.

[0064] In a seventh aspect, the present application further provides a computer program product, which, when executed, can implement the method described in the first aspect and any possible design thereof.

[0065] In an eighth aspect, the present application further provides a chip system, which includes one or more interface circuits and one or more processors; the interface circuits and the processors are interconnected by lines; the processors receive and execute computer instructions from the memory of the electronic device through the interface circuits to implement the method described in the first aspect and any possible design thereof.

[0066] In a ninth aspect, the present application further provides an artificial intelligence model, which has the function of implementing the method described in the first aspect and any possible design thereof.

[0067] Optionally, the artificial intelligence model includes one or more models. When the artificial intelligence model includes multiple models, the multiple models implement different functions in the method described in the first aspect and any possible design thereof.

[0068] In a tenth aspect, the present application further provides a communication system, which includes a first entity and a second entity, and the second entity interacts with the first entity to implement the method described in the first aspect and any possible design thereof.

[0069] It can be understood that for the beneficial effects that can be achieved by the sixth aspect to the tenth aspect provided above, reference can be made to the beneficial effects described in the first aspect to the fifth aspect, etc., and details are not repeated here.

[0070] It should be understood that the description of technical features, technical solutions, beneficial effects or similar languages in the present application does not imply that all features and advantages can be achieved in any single embodiment. On the contrary, it can be understood that the description of features or beneficial effects means that specific technical features, technical solutions or beneficial effects are included in at least one embodiment. Therefore, the description of technical features, technical solutions or beneficial effects in this specification does not necessarily refer to the same embodiment. Furthermore, the technical features, technical solutions and beneficial effects described in this embodiment can be combined in any appropriate manner. Those skilled in the art will understand that an embodiment can be implemented without one or more specific technical features, technical solutions or beneficial effects of a specific embodiment. In other embodiments, additional technical features and beneficial effects can also be identified in specific embodiments that do not embody all embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] Figure 1 Shows a schematic diagram of the composition of an electronic device provided by an embodiment of the present application;

[0072] Figure 2 It shows a schematic flowchart of the trust model evaluation method provided by an embodiment of the present application;

[0073] Figure 3 It shows a schematic diagram of the principle of a trust model evaluation system provided by an embodiment of the present application;

[0074] Figure 4 It shows another schematic flowchart of the trust model evaluation method provided by an embodiment of the present application;

[0075] Figure 5 It shows a schematic diagram of the principle of another trust model evaluation system provided by an embodiment of the present application;

[0076] Figure 6 It shows yet another schematic flowchart of the trust model evaluation method provided by an embodiment of the present application;

[0077] Figure 7 It shows yet another schematic flowchart of the trust model evaluation method provided by an embodiment of the present application;

[0078] Figure 8 It shows yet another schematic flowchart of the trust model evaluation method provided by an embodiment of the present application;

[0079] Figure 9 It shows yet another schematic flowchart of the trust model evaluation method provided by an embodiment of the present application;

[0080] Figure 10 It shows a schematic diagram of the composition of the trust model evaluation device provided by an embodiment of the present application. Detailed implementation manners

[0081] Trust assessment generally refers to evaluating the trust level of entities in an application scenario. By performing trust assessment on entities, it can be understood whether an entity is trustworthy, whether an entity can interact, or whether the data of an entity is available, etc.

[0082] For example, in some communication-related application scenarios, an entity can refer to an interaction object or a communication node in a communication network, such as a network device or a terminal device. According to the trust assessment result of a certain communication node in the communication network, it can be determined whether the communication node is secure, such as whether the data of the communication node is trustworthy, to detect the security and reliability of the communication network.

[0083] The method for trust evaluation may include: using a trust evaluation model to perform trust evaluation on an entity and outputting the trust evaluation result of the entity. Among them, the trust evaluation model is also called a trust model, and the trust evaluation result may refer to the trust degree of the entity. The trust degree may also be referred to as the degree of trust, or the credibility, or the reliability, etc. There is no limit to the description method of the trust degree here.

[0084] The granularity of the trust evaluation results output by different trust models may be different. For example, in some trust models, the trust degree of an entity is defined at two levels, including trustworthy or untrustworthy. The trust evaluation result of the entity output by the trust model is one of these two levels, such as the trust evaluation result is trustworthy or untrustworthy. The trust evaluation granularity of this type of trust model is relatively large (or relatively coarse).

[0085] There are also some trust models whose output trust evaluation results may be divided according to multiple levels of trust degree, or rather, the trust degree of an entity is defined at multiple levels, such as high, medium, and low. The trust evaluation result of the entity output by the trust model is one of the aforementioned multiple levels. The trust evaluation granularity of this type of trust model is related to the number of levels. The more levels there are, the smaller (or finer) the granularity is.

[0086] For an entity in a specific application scenario, currently, generally according to manual experience, a trust model is artificially selected to perform trust evaluation on the entity. However, the trust evaluation result output by the artificially selected trust model may have too large a granularity, resulting in low accuracy of the trust evaluation result, or it may also have too small a granularity, resulting in unnecessary computational cost increase.

[0087] Therefore, the embodiments of the present application provide a method for trust model evaluation, which can, for an entity in a specific application scenario, such as a first entity, evaluate the trust evaluation ability of the trust model according to the dimension of a first indicator, and provide a decision-making basis for the selection of the trust model.

[0088] For example, for a specific application scenario and an entity, it is possible to assist in selecting a trust model with a more appropriate trust evaluation granularity to perform trust evaluation on the entity in the application scenario, improve the accuracy of the trust evaluation result, and reduce unnecessary waste of computational cost.

[0089] Exemplarily, taking the trust evaluation of the first entity in the application scenario as an example, the method may include: performing trust evaluation on the first entity based on the trust model to obtain the trust evaluation result of the first entity, where the trust evaluation result includes at least one group, and each group of trust evaluation results corresponds to a first indicator, and the first indicator is used to indicate the evaluation dimension of the trust model; according to the trust evaluation result, evaluating the trust model to obtain the evaluation result of the trust model, and the evaluation result of the trust model is used to indicate the trust evaluation ability of the trust model within the dimension of the first indicator.

[0090] Exemplarily, the trust model evaluation method provided by the embodiments of the present application can be executed by an electronic device, or by a device built into the electronic device (e.g., a chip or one or more functional modules in the electronic device).

[0091] In some embodiments, the above-mentioned electronic device may be an entity in an application scenario or an electronic device related to the entity. For example, in a certain application scenario, when the first entity interacts with the second entity and the trust evaluation ability of the trust model is evaluated for this application scenario and the first entity, the above-mentioned electronic device may be the second entity itself (the second entity is an electronic device), or the above-mentioned electronic device may also be the device carrying the second entity or the electronic device included in the second entity, or the above-mentioned electronic device may further be another electronic device having data interaction or connection with the second entity.

[0092] In some other embodiments, the above-mentioned electronic device may be a device outside the application scenario. For example, the above-mentioned electronic device may be a computer or a server, or may also be other devices with data processing capabilities, such as mobile phones, computers, wireless terminals, etc.

[0093] It should also be understood that the above-mentioned electronic device may be a single electronic device or may be composed of multiple electronic devices. For example, taking a server as an example, the server may be a single server, or may also be a server cluster composed of multiple servers. In some implementation manners, the server cluster may also be a distributed cluster.

[0094] The present application does not limit the implementation manner or product form of the above-mentioned electronic device, nor the execution subject of the trust model evaluation method.

[0095] Exemplarily, Figure 1 shows a schematic diagram of the composition of an electronic device provided by the embodiments of the present application. In a possible implementation manner, the trust model evaluation method provided by the embodiments of the present application can be applied to Figure 1 the shown electronic device. As Figure 1 shown, the electronic device may include: at least one processor 11, a memory 12, a communication interface 13, and a bus 14.

[0096] The processor 11 is the control center of the electronic device and can be a single processor or a collective term for multiple processing elements. For example, the processor 11 can be a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present application, such as one or more digital signal processors (DSPs), or one or more field programmable gate arrays (FPGAs), etc.

[0097] Among them, the processor 11 can execute various functions of the electronic device by running or executing software programs stored in the memory 12 and calling data stored in the memory 12. For example, it can execute the steps included in the trust model evaluation method provided in the embodiments of the present application.

[0098] In a specific implementation, as an embodiment, the processor 11 can include one or more CPUs, such as Figure 1 the CPU0 and CPU1 shown in

[0099] In a specific implementation, as an embodiment, the electronic device can include multiple processors, such as Figure 1 the processor 11 and the processor 15 shown in

[0100] The memory 12 can store software programs of method steps executed by the electronic device and be controlled by the processor 11 for execution. The memory 12 can be a read-only memory (ROM) or other types of static storage devices that can store static information and instructions, a random access memory (RAM) or other types of dynamic storage devices that can store information and instructions, or can also be an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compact discs, laser discs, optical discs, digital versatile discs, Blu-ray discs, etc.), magnetic disk storage media or other magnetic storage devices, or any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto.

[0101] The memory 12 can exist independently and be connected to the processor 11 through the bus 14. Alternatively, the memory 12 can also be integrated with the processor 11, and this is not limited herein.

[0102] The communication interface 13 uses any device such as a transceiver for communicating with other devices or communication networks. The communication interface 13 can be an Ethernet interface, a radio access network (RAN) interface, a wireless local area networks (WLAN) interface, etc. The communication interface 13 can include a receiving unit to implement the receiving function and a transmitting unit to implement the transmitting function.

[0103] The bus 14 can be an industry standard architecture (ISA) bus, a peripheral component interconnect (PCI) bus, an extended industry standard architecture (EISA) bus, etc. This bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 1 only a thick line is used to represent it herein, but it does not mean that there is only one bus or one type of bus.

[0104] Although attached Figure 1The bus 14 is used in [the relevant context], but it can be understood that the bus can also be replaced by other forms of connection relationships, not limited to the bus itself.

[0105] It should be understood that Figure 1 The illustration shown is only for exemplary purposes. In the embodiments of the present application, the electronic device may also include more or fewer components than Figure 1 those shown, and no limitation is imposed herein.

[0106] The following provides an exemplary illustration of the trust model evaluation method provided by the embodiments of the present application.

[0107] It should be noted that the processing described below as being performed by a single execution entity can also be divided into being performed by multiple execution entities, and these execution entities can be logically and / or physically separated. For example, the functions involved in the trust model evaluation method can be divided into being performed by at least one functional module. Additionally, in the description of the embodiments of the present application, words such as "first", "second", etc. are only used for distinguishing descriptions and are not used for specifically limiting a certain feature. That is, the first or the second can include more content, rather than being limited to a specific concept. "And / or" describes the association relationship of associated objects and indicates that three relationships can exist. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. The character " / " generally represents an "or" relationship between the associated objects before and after. At least one means one or more; multiple means two or more. The embodiments of the present application can perform fewer steps than all the steps or more steps, without limitation. "At least one of the following" or its similar expressions are used to represent any combination of the items listed. For example, at least one of A, B, and (or) C can represent the following situations: A exists alone, B exists alone, C exists alone, A and B exist simultaneously, B and C exist simultaneously, A and C exist simultaneously, and A, B, and C exist simultaneously, where A, B, and C can be single or multiple.

[0108] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which this application belongs. The terms used in the specification of this application are only for the purpose of describing specific embodiments and are not intended to limit the present invention application.

[0109] As shown above, in the trust model evaluation method provided by the present application, the trust of the first entity can be evaluated based on the trust model to obtain the trust evaluation result of the first entity. The trust evaluation result includes at least one group, and each group of trust evaluation results corresponds to a first metric. The first metric is used to indicate the evaluation dimension of the trust model; according to the trust evaluation result, the trust model is evaluated to obtain the evaluation result of the trust model. The evaluation result of the trust model is used to indicate the trust evaluation ability of the trust model within the dimension of the first metric.

[0110] In some implementation manners, the trust evaluation result of the trust model for the first entity can be obtained by the user using the trust model to evaluate the first entity, or the trust evaluation result obtained from other existing data sources.

[0111] In other implementation manners, for a specific application scenario and entity, an evaluation environment can be constructed first, and the trust evaluation result of the trust model for the first entity can be obtained based on the evaluation environment. Taking this implementation manner as an example below, this method will be described.

[0112] For example, Figure 2 shows a schematic flowchart of the trust model evaluation method provided by an embodiment of the present application. As Figure 2 shown, the trust model evaluation method may include S201 - S203. Exemplarily, S201 - S203 can be executed by an electronic device or a device (such as a chip or one or more functional modules in the electronic device) built into the electronic device. The electronic device can refer to the electronic device described in the foregoing embodiments.

[0113] S201. Generate an evaluation environment according to the first metric, the scenario information of the application scenario, and the entity information of the first entity in the application scenario.

[0114] Among them, the evaluation environment can indicate the environment in which the first entity runs in the application scenario.

[0115] Exemplarily, in the embodiments of the present application, the application scenario may include a network security scenario, an intelligent system scenario, an online platform scenario, a complex decision-making environment scenario, etc. For any of the foregoing application scenarios, the first entity can be any real object or virtual object in the application scenario, such as a hardware device or module, an application or software module, data of hardware and / or software interaction, etc.

[0116] For example, in the network security scenario, with the increase in network attacks and data leakage incidents, it has become increasingly important to ensure the trustworthiness of network interactions and data transmissions. By evaluating the reliability of network nodes or data, potential security threats can be identified, ensuring data integrity and privacy. When the trust model evaluation method provided by the embodiments of this application is applied to the network security scenario, the first entity can be any object in a network (such as a communication network or other network), such as a network device, a hardware or software module running in the network device, data generated or interacted by the network device, etc.

[0117] For another example, intelligent systems can include an autonomous driving system, an intelligent medical decision-making system, etc., which are not limited here. In the autonomous driving system, the vehicle needs to evaluate the trustworthiness of other entities (such as pedestrians, other vehicles, or road signs, etc.) in the surrounding environment. The trust model can provide a trust rating for the vehicle based on the data and context of other entities, improving the safety of autonomous driving. When the trust model evaluation method provided by the embodiments of this application is applied to the autonomous driving system scenario, the first entity can be a pedestrian, a hardware or software module in the vehicle, other vehicles or road signs, or the data interacted or generated by these entities, etc. In the intelligent medical decision-making system, it is necessary to conduct trust evaluations on various medical devices, drugs, and treatment methods to improve safety. When the trust model evaluation method provided by the embodiments of this application is applied to the intelligent medical decision-making system scenario, the first entity can be a medical device, a drug, or a treatment method, etc.

[0118] For another example, online platforms can include e-commerce platforms, social networking platforms, etc., which are not limited here. In the e-commerce platform scenario, when making a purchase, the buyer needs to evaluate the reputation of the store or the product. The trust model can provide a more accurate and detailed trust evaluation for the user, enhancing their shopping experience. When the trust model evaluation method provided by the embodiments of this application is applied to the e-commerce platform scenario, the first entity can be the store or the product. In the social networking platform scenario, when a user interacts with unknown other users, the user hopes to know the credibility of the other users. The trust model can provide a detailed trust evaluation for them based on the historical behavior and feedback of the other users, helping the user better judge which other users to interact with. When the trust model evaluation method provided by the embodiments of this application is applied to the social networking platform scenario, the first entity can be the user account.

[0119] For another example, complex decision-making environments can include finance, insurance, etc. In the fields of finance and insurance, decisions often involve capital flows and risks, and decision-makers need to accurately evaluate the reliability of various investments or strategies. The trust model can flexibly adapt to different market conditions, dynamically evaluate various data and information, and provide clear and specific trust ratings for decision-makers. When the trust model evaluation method provided by the embodiments of this application is applied to the complex decision-making environment scenario, the first entity can be the data and information that need to be evaluated.

[0120] It should be understood that the above-described application scenarios are all for illustrative purposes, and this trust model evaluation method can also be applicable to more application scenarios with trust evaluation requirements. For example, this trust model evaluation method can be applicable to the Internet of Things scenario, where the first entity can be a device or a sensor, etc. Or, this trust model evaluation method can be applicable to the cloud computing scenario, where the first entity can be a computing device or a storage device, etc. Or, this trust model evaluation method can be applicable to the edge computing scenario, where the first entity can be an edge node or a central node. Or, this trust model evaluation method can be applicable to network slicing and virtualized network scenarios, software-defined network scenarios, blockchain scenarios, etc. This application will not list them one by one, but this application does not limit the application scenarios to which this trust model evaluation method can be applicable.

[0121] In S201, the scenario information of the application scenario can be used to describe the content of the application scenario. For example, the scenario information can include which entities are included in the application scenario, the connection or interaction relationship between the entities, etc. The first entity can be any one or more entities in the application scenario. The entity information of the first entity can be used to describe the first entity. For example, the entity information of the first entity can include the static attribute information of the first entity and / or the dynamic behavior information. The static attribute information of the first entity can include the attribute information of the first entity itself. The dynamic behavior information of the first entity can include the relevant information of the first entity's operation or running, the interaction information with other entities, etc.

[0122] Exemplarily, taking the application scenario as a communication network, such as a cellular network or an Internet Protocol (IP) network, the first entity can be a network element in the communication network, such as an access network device (such as a base station), a terminal device, a core network device, etc. The scenario information can be which network elements are included in the communication network, the connection relationship and interaction relationship between the network elements, etc. The entity information of the first entity can be the information such as the packet loss rate of the first network element and whether the packet is forwarded along the path.

[0123] It can be understood that before S201, this method can also include the step of obtaining the scenario information of the application scenario and the entity information of the first entity in the application scenario.

[0124] In some possible implementations, for a specific application scenario and a first entity, the scenario information of the application scenario and the entity information of the first entity can be obtained from a specified data source and / or interface. For example, for an autonomous driving system, the scenario information of the autonomous driving system and the entity information of the first entity (such as a vehicle) in the autonomous driving system can be obtained from the application programming interface (API) provided by the autonomous driving system. Another example is that for a cellular network, the scenario information of the cellular network and the entity information of the first entity (such as a first network element) in the cellular network can be obtained from the southbound / northbound interface of the cellular network, the serial bus interface (SBI), etc.

[0125] In some other possible implementations, the scenario information of the application scenario and the entity information of the first entity can also be input by a user. For example, a human-machine interaction interface can be provided, and the user can input the scenario information of the application scenario and the entity information of the first entity through the human-machine interaction interface.

[0126] After obtaining the scenario information of the application scenario and the entity information of the first entity in the application scenario, S201 can be executed. In S201, the first metric can be understood as a dimension for evaluating the trust model. The first metric can include at least one, that is, the trust model can be evaluated according to one or more dimensions.

[0127] For example, in a possible design, the first metric can include one or more of the following metrics: security of data, comprehensiveness of the trust evaluation result, usability of the trust model, functionality of the trust model, robustness of the trust model, neutrality of the trust evaluation result, explicability of the trust evaluation result. These are hereinafter simply referred to as security, comprehensiveness, usability, functionality, robustness, neutrality, and explicability in turn.

[0128] Among them, security can be used to evaluate whether the trust model can ensure the security of data when processing data and making predictions, preventing data leakage or malicious exploitation. Comprehensiveness can be used to evaluate whether the trust model can cover relevant factors and variables to make predictions and analyses as comprehensively as possible. Availability can be used to evaluate whether the trust model can be effectively invoked and used when needed and can maintain stable and reliable performance in different environments. Functionality can be used to evaluate whether the trust model can effectively implement its designed functions, including data processing, feature extraction, prediction, and analysis, etc. Robustness can be used to evaluate whether the trust model can maintain stable performance in the face of abnormal or noisy data and will not show serious deviations or errors due to data changes. Neutrality can be used to evaluate whether the trust model can maintain an objective and neutral stance when making predictions and analyses without being affected by subjective factors. Interpretability can be used to evaluate whether the trust assessment can clearly explain its prediction results and analysis process, enabling users to understand the decision-making basis and reasoning process of the model.

[0129] In S201, an evaluation environment can be generated according to the first metric, the scenario information of the application scenario, and the entity information of the first entity in the application scenario. The application scenario can be restored in the evaluation environment. In the embodiments of the present application, the evaluation environment can be understood as a test environment, and the performance of the trust model within each dimension of the first metric can be tested in the evaluation environment, or rather, the performance shown by the trust model in each dimension corresponding to the first metric, such as the trust assessment ability of the trust model within the dimension of the first metric.

[0130] Exemplarily, the evaluation environment can include an environment module and an agents module. The environment module can provide environmental information such as the scenario information of the application scenario and the entity information of the first entity, and the agents module can invoke the trust model to be evaluated and evaluate the trust degree of the first entity based on the environmental information, and output the trust assessment result of the first entity. For example, S202 can be executed.

[0131] S202: Invoke the trust model, conduct a trust assessment on the first entity based on the evaluation environment, and output the trust assessment result of the first entity. The first metric includes at least one metric, and the trust assessment result of the first entity includes the trust assessment result corresponding to each first metric.

[0132] Alternatively, S202 can also be described as: Invoke the trust model, conduct a trust assessment on the first entity based on the evaluation environment, and obtain the trust assessment result of the trust model for the first entity.

[0133] As described in the foregoing embodiments, the granularity of the trust evaluation results output by different trust models may vary. In S202, each or any granularity level of the trust model can be invoked to output the trust evaluation result of the first entity, so as to evaluate the trust evaluation ability of the trust model at this granularity level.

[0134] For example, the trust model described in S202 can be any of the following granularity levels: binary level, coarse-grained level, and fine-grained level.

[0135] Among them, the binary level trust model can be used to evaluate whether an entity can be trusted. Its trust evaluation results have two states: can be trusted or cannot be trusted. For example, an entity can be marked as 1 (trusted) or 0 (not trusted). The decision is clear, relatively simple, without ambiguity and easy to interpret. There is no need for detailed measurement or calculation of trust. The computing resources and time required during the evaluation process are less. It can be used in scenarios that require quick decisions, such as binary authentication systems, quick screening, etc.

[0136] The granularity of the coarse-grained level trust model is finer than that of the binary level trust model, or rather, the level of the trust evaluation results of the coarse-grained level trust model is more than that of the binary level trust model. The coarse-grained level trust model defines several preset classification levels for the degree of trust, such as "highly trusted", "moderately trusted", "lowly trusted", etc., providing greater flexibility than the binary level trust model. The coarse-grained level trust model can be applied to some scenarios where it is necessary to know the general trust level of an entity without knowing its specific trust value. For example, in some recommendation systems, users may only care about whether they should "highly trust", "moderately trust" or "lowly trust" a certain recommendation, rather than knowing the specific degree of trust or trust score.

[0137] The granularity of the fine-grained level trust model is finer than that of the coarse-grained level trust model, and it can capture more detailed trust information. For example, continuous numerical values or more classification levels can be used to represent the degree of trust. The fine-grained level trust model can be applied to scenarios that require highly accurate and detailed trust evaluation, such as complex decision support systems, security evaluations, detailed user feedback systems, etc.

[0138] Exemplarily, in S202, environmental information such as the scenario information of the application scenario and the entity information of the first entity can be input into the trust model. The trust model performs a trust assessment on the first entity based on the scenario information of the application scenario and the entity information of the first entity, and outputs the trust assessment result of the first entity. Among them, the trust assessment result of the first entity is related to the first indicator, and may include the trust assessment result corresponding to each first indicator. The trust assessment result corresponding to one first indicator can be referred to as a group, that is, each group of trust assessment results corresponds to one first indicator, and the first indicators corresponding to different groups of trust assessment results may be different.

[0139] For example, taking the first indicator including robustness as an example, for robustness, some abnormal data or noise data can be added to the input data, and the first entity is subjected to multiple trust assessments to obtain multiple trust assessment results (that is, the trust assessment result of the first entity corresponding to robustness). The foregoing multiple trust assessment results can reflect the robustness of the trust model and can indicate that the trust model maintains stable performance in the face of abnormal or noise data.

[0140] Another example, taking the first indicator including interpretability as an example, for interpretability, it can be determined whether the interpretability of the trust model matches the application scenario and the first entity by analyzing whether the trust assessment result of the first entity meets the assessment requirements. Any one trust assessment result can be used as the trust assessment result of the first entity corresponding to interpretability.

[0141] It should be understood that for different first indicators, the corresponding trust assessment results of the first entity may be the same or different. For example, the trust assessment results of the first entity corresponding to functionality and interpretability may be the same. Or, the trust assessment result of the first entity corresponding to interpretability is the trust assessment result of one output, and the trust assessment result of the first entity corresponding to robustness is the trust assessment result of multiple outputs, etc.

[0142] The process described in S201 - S202 above can be understood as a process of performing a trust assessment on the first entity based on the trust model to obtain the trust assessment result of the first entity. For example, performing a trust assessment on the first entity based on the trust model to obtain the trust assessment result of the first entity includes: obtaining the first indicator, the scenario information of the application scenario, and the entity information of the first entity; generating an assessment environment according to the first indicator, the scenario information of the application scenario, and the entity information of the first entity, where the assessment environment indicates the environment in which the first entity runs in the application scenario; calling the trust model, and performing a trust assessment on the first entity based on the assessment environment to obtain the trust assessment result of the trust model for the first entity.

[0143] Among them, the trust assessment result includes the trust assessment results corresponding to at least one first indicator of the trust model respectively, and the first indicator is used to indicate the evaluation dimension of the trust model.

[0144] After obtaining the trust evaluation result corresponding to the first metric, S203 can be executed to evaluate the trust evaluation ability of the trust model.

[0145] S203. Output / determine the evaluation result of the trust model according to the trust evaluation result corresponding to the first metric. The evaluation result of the trust model is used to indicate the trust evaluation ability of the trust model within the dimension of the first metric.

[0146] Alternatively, S203 can also be described as: evaluating the trust model according to the trust evaluation result to obtain the evaluation result of the trust model.

[0147] Exemplarily, for each first metric, the performance of the trust model within the dimension of the first metric can be analyzed based on the trust evaluation result corresponding to the first metric, or in other words, the performance of the trust model on the first metric. This performance can be referred to as the trust evaluation ability of the trust model. Among them, the analysis result can be called the evaluation result of the trust model, and the evaluation result of the trust model can indicate the trust evaluation ability of the trust model within the dimension of the first metric.

[0148] For example, taking the first metric as interpretability, that is, the trust evaluation result of the first entity includes the trust evaluation result corresponding to interpretability as an example. In S203, the performance of the trust model in the dimension of interpretability can be analyzed and evaluated according to the trust evaluation result corresponding to interpretability, and the evaluation result of the trust model regarding interpretability can be obtained. This evaluation result can be used to indicate the trust evaluation ability of the trust model within the dimension of interpretability. For example, it can indicate whether the trust model can clearly explain its prediction results and analysis process, and whether the granularity of the trust evaluation result output by the trust model is appropriate, etc.

[0149] Optionally, the evaluation result of the trust model regarding interpretability can be output by using deep learning or machine learning methods according to the trust evaluation result corresponding to interpretability. For example, the features of the trust evaluation result (such as granularity, user feedback, information entropy, etc.) can be used as inputs, and the annotation information of the trust evaluation result, such as scoring (the score is used to indicate the appropriateness of the trust evaluation granularity) or labels (such as appropriate or inappropriate) can be used as outputs. A neural network model can be pre-trained in advance. This neural network model has the function of outputting whether the trust evaluation granularity is appropriate or the degree of appropriateness (such as a score) according to the input features of the trust evaluation result. Inputting the aforementioned trust evaluation result corresponding to interpretability into this neural network model, the neural network model can output the evaluation result of the trust model regarding interpretability, such as whether the trust evaluation granularity is appropriate.

[0150] Similarly, for other first metrics, the steps of outputting the evaluation result of the trust model according to the trust evaluation result corresponding to the first metric as described above can also be implemented by means of deep learning or machine learning through a neural network model. Examples are not given one by one here. It should be understood that for different first metrics, the steps of outputting the evaluation result of the trust model according to the trust evaluation result corresponding to the first metric as described above can be implemented by the same or different neural network models.

[0151] As described above, the trust model evaluation method provided by the embodiments of the present application can evaluate the trust evaluation ability of the trust model for an entity in a specific application scenario, such as a first entity, according to the dimension of the first metric. By evaluating the trust evaluation ability of the trust model, it is possible to assist in selecting a more suitable trust model for the trust evaluation of the entity in the application scenario, or to assist the user in judging whether the selected trust model is appropriate. Or rather, it can provide a decision-making basis for the selection of the trust model. For example, the user can judge whether the trust evaluation granularity is appropriate according to the evaluation result of the trust model, or select a trust model with a more appropriate trust evaluation granularity.

[0152] For example, taking the above-mentioned binary level trust model, coarse-grained level trust model, and fine-grained level trust model as examples, the binary level trust model has a relatively coarse trust evaluation granularity, a simple evaluation method, and low computational resource consumption. However, it lacks consideration of the intermediate state or continuous level of the trust degree. Its black-and-white nature makes it easy to ignore subtle differences, which may lead to misjudgments. The binary level trust model may ignore subtle differences and evaluate objects at different trust levels as both "good" or both "bad", and it is unable to further compare the "good" models. The binary level trust model handles edge cases relatively roughly.

[0153] The fine-grained level trust model can provide in-depth and detailed trust evaluations, but it is relatively complex and may consume relatively more computational resources.

[0154] Compared with the binary level trust model and the fine-grained level trust model, the coarse-grained level trust model provides a tool for applications that require a certain degree of classification or level. It is more flexible than the binary level trust model but not as complex as the fine-grained level trust model. The coarse-grained level trust model can provide a balance between simplicity and detailed measurement, providing a tool for users that is neither overly simplified nor overly complex. However, the coarse-grained level trust model may ignore subtle differences, may evaluate objects with different trust levels as the same class of hierarchical categories, and may also have errors in dealing with edge cases to a certain extent.

[0155] When conducting trust assessment on entities in specific application scenarios, for some application scenarios, it may be more appropriate to use the binary level trust model for quick judgment. For some other application scenarios, it may be more appropriate to use the fine-grained level trust model for fine-grained assessment. And for some other application scenarios, it may be more appropriate to use the coarse-grained level trust model for relatively coarse-grained assessment without wasting too much computing resources. For users, when manually selecting a trust model to conduct trust assessment on entities, there may be problems such as the trust assessment result output by the selected trust model having too large a granularity, resulting in low accuracy of the trust assessment result, or having too small a granularity, resulting in increased unnecessary computing costs. However, through the trust model assessment method provided by the embodiments of the present application, the trust assessment ability of the trust model can be evaluated to assist or help users select a more suitable trust model for the trust assessment of entities in the application scenario. For example, users can judge whether the trust assessment granularity is appropriate based on the assessment result of the trust model, or select a trust model with a more appropriate trust assessment granularity.

[0156] In addition, in this trust model assessment method, for a specific application scenario and the first entity in the application scenario, after generating an assessment environment, the trust assessment ability of the trust model is further evaluated based on the assessment environment, which can enable the assessment environment to be flexibly adjusted according to the application scenario and the first entity. This method can be more flexibly applicable to the assessment of trust models in more application scenarios. In other words, the implementation framework of this method can be self-flexibly adjusted according to the changes in the application scenario and the first entity, and has strong adaptability.

[0157] It should also be understood that in the embodiments of the present application, the richer the categories of the first indicators are, the more comprehensive the result of evaluating the trust evaluation ability of the trust model will be. According to the trust evaluation results corresponding to different first indicators, the evaluation result of the trust model is output, which can enable the evaluation result of the trust model to reflect the trust evaluation ability of the trust model in different first indicator dimensions, and display or describe the performance of the trust model for users in different first indicator dimensions, or rather, more comprehensively display the comprehensive performance of the trust model. For example, the embodiments of the present application can provide feedback to the developers and users of the trust model within different first indicator dimensions, enabling them to more clearly understand the performance of the trust model in different dimensions.

[0158] In the above embodiments, the trust model examples are divided from the perspective of trust evaluation granularity. Optionally, the trust model can also be classified from the perspective of function, and the trust model corresponding to each function can be further divided according to the evaluation granularity. For different application scenarios, the trust models may be the same or different. For example, in a communication network, for a peer-to-peer file sharing and communication environment, considering the characteristics of direct interaction between users and emphasizing the behavior evaluation of peers, a point-to-point (P2P) trust model can be adopted. For evaluating a traditional computer network environment, with the focus on the trustworthiness of devices and data, a network trust model can be adopted. For a mobile Ad hoc wireless network, dealing with a dynamic and decentralized network environment, an Ad hoc trust model can be adopted. For a wireless sensor network, emphasizing the problems of resource limitation and energy efficiency, a wireless sensor network (WSN) trust model can be adopted, etc.

[0159] In a possible design, before performing the above S201, the scenario information of the application scenario and / or the entity information of the first entity in the application scenario can also be preprocessed. For example, noise and irregular terms in the scenario information and entity information can be removed, such as removing abnormal data and filling in missing data. The specific method of preprocessing is not limited here.

[0160] In this design, by preprocessing the scenario information of the application scenario and / or the entity information of the first entity in the application scenario, the accuracy of the scenario information and entity information can be improved, thereby generating a better or more stable evaluation environment to improve the accuracy of the trust model evaluation.

[0161] In a possible design, the step of generating an evaluation environment according to the first metric, the scenario information of the application scenario, and the entity information of the first entity in the application scenario in S201 described above may include: extracting the scenario features of the scenario information of the application scenario and the entity features of the entity information of the first entity; generating an evaluation environment according to the first metric, the scenario features, and the entity features of the first entity.

[0162] Exemplarily, a feature extraction model trained by machine learning or deep learning may be used to extract the scenario features of the scenario information of the application scenario and the entity features of the entity information of the first entity. The training method of the feature extraction model is relatively mature and will not be elaborated here.

[0163] In this design, by extracting the scenario features of the scenario information of the application scenario and the entity features of the entity information of the first entity, and generating an evaluation environment according to the first metric, the scenario information, and the entity features of the first entity, the effectiveness or accuracy of the evaluation environment can be further improved, the application scenario can be restored more accurately, and thus the accuracy of the trust model evaluation can be improved.

[0164] Optionally, generating an evaluation environment according to the first metric, the scenario features, and the entity features of the first entity may include: fusing the scenario features and the entity features of the first entity to obtain a fused feature; generating an evaluation environment according to the first metric and the fused feature. Fusing the scenario features and the entity features of the first entity first and then generating an evaluation environment according to the obtained fused feature can improve the authenticity and accuracy of the evaluation environment.

[0165] Exemplarily, the trust model evaluation method provided in the embodiments of the present application may be implemented by a trust model evaluation system. For example, Figure 3 shows a schematic diagram of the principle of a trust model evaluation system provided in the embodiments of the present application. As Figure 3 shown, the trust model evaluation system may include: an entity analysis module 311, a scenario analysis module 312, a fusion module 320, an evaluation module 330, a metric storage module 340, a trust model 350, and a parsing module 360. Each of the foregoing modules may be a software module, and these software modules may be integrated on one device or deployed on different devices, which is not limited herein.

[0166] Among them, the entity analysis module 311 may obtain the entity information of the first entity, extract the entity features of the first entity according to the entity information of the first entity, and transmit the entity features of the first entity to the fusion module 320. The scenario analysis module 312 may obtain the scenario information of the application scenario, extract the scenario features of the application scenario according to the scenario information of the application scenario, and transmit the scenario features of the application scenario to the fusion module 320.

[0167] The fusion module 320 can fuse the scenario features of the application scenario and the entity features of the first entity, and generate an evaluation environment based on the first metric and the fusion result. For example, the fusion result can be referred to as the fusion feature. The fusion module 320 can transmit the information of the generated evaluation environment to the evaluation module 330. The first metric can be stored in the metric storage module 340 and provided to the fusion module 320 by the metric storage module 340.

[0168] The evaluation module 330 can call the trust model 350 to perform a trust evaluation on the first entity based on the evaluation environment, and output the trust evaluation result of the first entity. The trust evaluation result of the first entity includes the trust evaluation result corresponding to each first metric.

[0169] The evaluation module 330 can transmit the trust evaluation result corresponding to the first metric to the parsing module 360. The parsing module 360 can output the evaluation result of the trust model (or referred to as the model evaluation result) according to the trust evaluation result corresponding to the first metric.

[0170] The principle of the above-mentioned trust model evaluation system can also be referred to as follows Figure 4 . Figure 4 Fig. shows another schematic flowchart of the implementation process of the trust model evaluation method provided by the embodiment of the present application. As Figure 4 shown, based on Figure 3 the trust model evaluation system shown, the trust model evaluation method can include S401-S411.

[0171] S401. The interface module receives the application scenario and the first entity input by the user.

[0172] Exemplarily, the trust model evaluation system may further include an interface module ( Figure 3 not shown). The interface module can provide an input interface for the user. The user can input the specific application scenario and the first entity that need to be evaluated by the trust model, such as inputting the scenario name and the entity name. The interface module can obtain the scenario information of the application scenario and the entity information of the first entity according to the application scenario and the first entity input by the user. The obtaining method can refer to that described in the foregoing embodiments and will not be elaborated here.

[0173] S402. The interface module sends the scenario information of the application scenario and the entity information of the first entity to the analyzing module.

[0174] Among them, the analyzing module may include Figure 3The entity analyzing module and scenario analyzing module shown in []. The interface module can send the entity information of the first entity to the entity analyzing module and the scenario information of the application scenario to the scenario analyzing module.

[0175] Correspondingly, the analyzing module receives the scenario information of the application scenario and the entity information of the first entity.

[0176] Optionally, the interface module can send an assessment request message to the analyzing module, and the assessment request message includes the scenario information of the application scenario and the entity information of the first entity.

[0177] S403. The analyzing module extracts the entity features of the first entity and the scenario features of the application scenario.

[0178] S404. The analyzing module sends the entity features of the first entity and the scenario features of the application scenario to the fusion module.

[0179] Correspondingly, the fusion module receives the entity features of the first entity and the scenario features of the application scenario.

[0180] S405. The fusion module obtains the first metric from the metric storage module.

[0181] Or rather, the metric storage module sends the first metric to the fusion module, and the fusion module receives the first metric.

[0182] S406. The fusion module generates an evaluation environment according to the first metric, the entity features of the first entity, and the scenario features of the application scenario.

[0183] S407. The fusion module sends the information of the evaluation environment to the assessment module.

[0184] Correspondingly, the assessment module receives the information of the evaluation environment.

[0185] S408. The assessment module calls the trust model to perform a trust assessment on the first entity based on the evaluation environment.

[0186] S409. The trust model returns the trust assessment result corresponding to the first metric to the assessment module.

[0187] That is to say, the trust model outputs the trust assessment result of the first entity, and the trust assessment result of the first entity includes the trust assessment result corresponding to each first metric.

[0188] S410. The evaluation module sends the trust evaluation result corresponding to the first metric to the parsing module.

[0189] Correspondingly, the parsing module receives the trust evaluation result corresponding to the first metric.

[0190] S411. The parsing module outputs the evaluation result of the trust model according to the trust evaluation result corresponding to the first metric.

[0191] Exemplarily, the parsing module can store and parse the trust evaluation result to obtain the evaluation result of the trust model.

[0192] Optionally, the first metric used to generate the evaluation environment in this trust model evaluation method is related to the application scenario and / or the first entity. For example, the method further includes: outputting / determining the first metric according to the scenario information and the entity information of the first entity.

[0193] Exemplarily, the steps of obtaining the first metric, the scenario information of the application scenario, and the entity information of the first entity in the foregoing embodiments may include: obtaining the scenario information of the application scenario and the entity information of the first entity; based on the scenario information of the application scenario and the entity information of the first entity, obtaining the first metric.

[0194] In a possible design, the first metric corresponding to the specific application scenario and the first entity may be predefined, or preconfigured, or configured. For example, evaluation metrics matching the application scenario and the entity can be configured to obtain the mapping relationship between the evaluation metrics, the application scenario, and the entity. Based on the scenario information of the application scenario and the entity information of the first entity, obtaining the first metric may include: determining the evaluation metric matching the application scenario and the first entity according to this mapping relationship, such as a target evaluation metric, and selecting at least one metric from the target evaluation metrics as the first metric. Or rather, the mapping relationship may be predefined, or preconfigured, or configured.

[0195] For example, in one implementation, the mapping relationship can be preconfigured in the hardware and / or software of the above electronic device, such as being previously recorded / written into the above metric storage module and can be changed through software or hardware.

[0196] For another example, in another implementation, the mapping relationship can be configured to the above electronic device in a configured manner, such as being recorded / written into the hardware and / or software of the above electronic device.

[0197] For yet another example, in yet another implementation, the mapping relationship does not require configuration by other devices and can be predefined (can be previously recorded / written) information in the hardware and / or software of the above electronic device, or can be understood as information that cannot be changed by other devices.

[0198] This application places no restrictions on the implementation manner of the mapping relationship.

[0199] Exemplarily, developers or other users can pre-define or configure the application scenarios and the mapping relationship or association relationship between the first entity and the first metric. For different application scenarios and / or the first entity, the corresponding first metric can be the same or different.

[0200] This design can implement the customization of selecting the first metric for the current application environment and the first entity according to the user design, and evaluating the trust model, which can improve the accuracy of the trust model evaluation result, so that the trust model evaluation result reflects the performance of the trust model from a more accurate and effective dimension.

[0201] In another possible design, the first metric used to generate the evaluation environment can also be dynamically determined according to the scenario information and the entity information of the first entity. For example, based on the scenario information of the application scenario and the entity information of the first entity, obtaining the first metric may include: through deep learning or machine learning, a pre-trained neural network model dynamically determines an evaluation metric that matches the application scenario and the first entity according to the scenario information and the entity information of the first entity, such as a so-called target evaluation metric, and selects at least one metric from the target evaluation metrics as the first metric. For example, the scenario information and the entity information of the first entity (or scenario features and entity features, or fused features) can be input into a pre-trained neural network model, and the pre-trained neural network model can output the first metric that matches the application scenario and the first entity according to the input.

[0202] Exemplarily, obtaining the first metric based on the scenario information of the application scenario and the entity information of the first entity includes: fusing the scenario information of the application scenario and the entity information of the first entity to obtain a fused feature; obtaining the first metric based on the fused feature. The fused feature can be specifically seen in the foregoing embodiments. For example, the scenario features of the scenario information of the application scenario and the entity features of the entity information of the first entity can be extracted, and the scenario features and the entity features are fused to obtain a fused feature.

[0203] Obtaining the first metric based on the fused feature can further improve the effectiveness or accuracy of the first metric, and select a more suitable first metric for the application scenario and the first entity to evaluate the trust model.

[0204] It can be understood that the neural network model can be trained by using the scenario features and the entity features as inputs and the first metric labeled for the scenario features and the entity features as outputs. The training process will not be elaborated here.

[0205] Exemplarily, Figure 5The schematic diagram of the principle of another trust model evaluation system provided by the embodiments of the present application is shown. As Figure 5 shown, on the basis of the above Figure 3 shown, the trust model evaluation system may further include: an index generation module 370. The index generation module 370 may be a software module.

[0206] Among them, the fusion module 320 may also transmit the fusion result of the scenario feature and the entity feature (such as the fusion feature) to the index generation module 370. The index generation module 370 may adopt a pre-trained neural network model and output a first index matching the application scenario and the first entity according to the fusion feature. The index generation module 370 may transmit the generated first index to the index storage module 340 for storage. The index storage module 340 may provide the first index matching the application scenario and the first entity to the fusion module 320 for the fusion module 320 to generate an evaluation environment.

[0207] Figure 5 The principle of the trust model evaluation system described above may also refer to the following Figure 6 . Figure 6 The schematic diagram of another implementation process of the trust model evaluation method provided by the embodiments of the present application is shown. As Figure 6 shown, based on the Figure 5 shown trust model evaluation system, the trust model evaluation method may include S601 - S615.

[0208] S601. The interface module receives the application scenario and the first entity input by the user.

[0209] S602. The interface module sends the scenario information of the application scenario and the entity information of the first entity to the analysis module.

[0210] Correspondingly, the analysis module receives the scenario information of the application scenario and the entity information of the first entity.

[0211] S603. The analysis module extracts the entity feature of the first entity and the scenario feature of the application scenario.

[0212] S604. The analysis module sends the entity feature of the first entity and the scenario feature of the application scenario to the fusion module.

[0213] Correspondingly, the fusion module receives the entity feature of the first entity and the scenario feature of the application scenario.

[0214] S601 - S604 may refer to the above S401 - S404 and will not be elaborated here.

[0215] S605. The fusion module fuses the entity feature of the first entity and the scenario feature of the application scenario to obtain a fusion feature.

[0216] S605 can refer to the above-mentioned fusion of entity features and scenario features in S406, which will not be elaborated here.

[0217] S606. The fusion module sends the fused features to the metric generate module.

[0218] Correspondingly, the metric generate module receives the fused features.

[0219] S607. The metric generate module outputs a first metric that matches the application scenario and the first entity based on the fused features.

[0220] S606 - S607 can refer to the above-mentioned method of dynamically determining the first metric by the pre-trained neural network model, which will not be elaborated here.

[0221] S608. The metric generate module sends the first metric that matches the application scenario and the first entity to the metric storage module.

[0222] Correspondingly, the metric storage module receives the first metric that matches the application scenario and the first entity.

[0223] S609. The fusion module obtains the first metric from the metric storage module.

[0224] It can be understood that the first metric is the metric that matches the application scenario and the first entity.

[0225] S610. The fusion module generates an evaluation environment based on the first metric, the entity features of the first entity, and the scenario features of the application scenario.

[0226] S611. The fusion module sends the information of the evaluation environment to the evaluation module.

[0227] Correspondingly, the evaluation module receives the information of the evaluation environment.

[0228] S612. The evaluation module calls the trust model to perform a trust evaluation on the first entity based on the evaluation environment.

[0229] S613. The trust model returns the trust evaluation result corresponding to the first metric to the evaluation module.

[0230] S614. The evaluation module sends the trust evaluation result corresponding to the first metric to the parsing module.

[0231] Correspondingly, the parsing module receives the trust evaluation result corresponding to the first metric.

[0232] S615. The parsing module outputs the evaluation result of the trust model according to the trust evaluation result corresponding to the first metric.

[0233] S609 - S615 can refer to the above - described S405 - S411 and will not be elaborated here.

[0234] This design can use deep learning or machine learning to dynamically and specifically select the first metric for the current application environment and the first entity in a customized manner to evaluate the trust model, which can improve the accuracy of the trust model evaluation results, enhance the adaptability of the trust model evaluation for different application scenarios, increase the flexibility of the evaluation, and enable the trust model evaluation results to reflect the performance of the trust model from a more accurate and effective dimension.

[0235] In some other possible designs, the first metric used to generate the evaluation environment in this trust model evaluation method can also be one or more first metrics selected by the user from the first metrics obtained based on the scenario information of the application scenario and the entity information of the first entity. For example, the user can perform an operation to select the first metric, and the method further includes: in response to the user's selection operation of the first metric, using the first metric indicated by the selection operation as the first metric for generating the evaluation environment. In this design, the first metric obtained based on the scenario information of the application scenario and the entity information of the first entity can be referred to as the alternative first metric, and the first metric selected by the user from the alternative first metrics can be referred to as the target first metric.

[0236] For example, the step of generating the evaluation environment according to the first metric, the scenario information of the application scenario, and the entity information of the first entity in the application scenario described in the above - mentioned embodiments can include: in response to the user's selection operation of the first metric (alternative first metric), generating the evaluation environment according to the first metric (target first metric) indicated by the selection operation, the scenario information, and the entity information of the first entity.

[0237] Exemplarily, taking the function of enabling the user to select the first metric in the process of the trust model evaluation method based on the above - mentioned Figure 6 shown as an example, Figure 7 Fig. shows another schematic flowchart of the trust model evaluation method provided by an embodiment of the present application. As Figure 7 shown, the trust model evaluation method can include S701 - S717.

[0238] S701. The interface module receives the application scenario and the first entity input by the user.

[0239] S702. The interface module sends the scenario information of the application scenario and the entity information of the first entity to the analysis module.

[0240] Correspondingly, the analysis module receives the scenario information of the application scenario and the entity information of the first entity.

[0241] S703. The analysis module extracts the entity features of the first entity and the scenario features of the application scenario.

[0242] S704. The analysis module sends the entity features of the first entity and the scenario features of the application scenario to the fusion module.

[0243] Correspondingly, the fusion module receives the entity features of the first entity and the scenario features of the application scenario.

[0244] S705. The fusion module fuses the entity features of the first entity and the scenario features of the application scenario to obtain fused features.

[0245] S706. The fusion module sends the fused features to the metric generation module.

[0246] Correspondingly, the metric generation module receives the fused features.

[0247] S707. The metric generation module outputs a first metric that matches the application scenario and the first entity according to the fused features.

[0248] S708. The metric generation module sends the first metric that matches the application scenario and the first entity to the metric storage module.

[0249] Correspondingly, the metric storage module receives the first metric that matches the application scenario and the first entity.

[0250] S701 - S708 can be referred to the above S601 - S608 and will not be elaborated here.

[0251] S709. The metric storage module sends the first metric that matches the application scenario and the first entity to the interface module.

[0252] Correspondingly, the interface module receives the first metric that matches the application scenario and the first entity.

[0253] After the interface module receives the first metric that matches the application scenario and the first entity (i.e., the alternative first metric), it can display the first metric that matches the application scenario and the first entity to the user for selection. The user can select one or more target first metrics from the alternative first metrics. For example, the interface module can display the alternative first metrics to the user through a human - machine interaction interface, and the user can perform a selection operation on the human - machine interaction interface to select the target first metric. The implementation method of the user's selection operation is not limited here.

[0254] After the user selects the target first metric, the interface module can execute S710.

[0255] S710. The interface module sends the first metric indicated by the selection operation, that is, the target first metric, to the metric storage module.

[0256] Accordingly, the metric storage module receives the first metric indicated by the selection operation.

[0257] Exemplarily, after the metric storage module receives the first metric indicated by the selection operation, it can update the stored first metric that matches the application scenario and the first entity to the first metric indicated by the selection operation.

[0258] S711. The fusion module obtains the first metric indicated by the selection operation from the metric storage module.

[0259] S712. The fusion module generates an evaluation environment according to the first metric indicated by the selection operation, the entity feature of the first entity, and the scenario feature of the application scenario.

[0260] S713. The fusion module sends the information of the evaluation environment to the evaluation module.

[0261] Accordingly, the evaluation module receives the information of the evaluation environment.

[0262] S714. The evaluation module invokes the trust model to perform a trust evaluation on the first entity based on the evaluation environment.

[0263] S715. The trust model returns the trust evaluation result corresponding to the first metric indicated by the selection operation to the evaluation module.

[0264] S716. The evaluation module sends the trust evaluation result corresponding to the first metric indicated by the selection operation to the parsing module.

[0265] Accordingly, the parsing module receives the trust evaluation result corresponding to the first metric indicated by the selection operation.

[0266] S717. The parsing module outputs the evaluation result of the trust model according to the trust evaluation result corresponding to the first metric indicated by the selection operation.

[0267] S711 - S717 can be referred to the above S609 - S615 and will not be elaborated here.

[0268] In this design, while realizing the customization of selectively choosing the first metric for the current application environment and the first entity, it also allows users to adjust the first metric according to their specific needs. For example, users can select more meaningful or relevant first metrics according to their needs, which can further improve the value of the evaluation result of the trust model to users. In addition, by screening and selecting more relevant first metrics, it is also possible to reduce the noise brought by irrelevant or secondary metrics and further improve the accuracy of the evaluation result.

[0269] Alternatively, users can also select some specific metrics for evaluation according to the need to save evaluation time and resources, so as to improve the evaluation efficiency.

[0270] It should be understood that the types of the first indicators described in the above embodiments are all for illustrative purposes. In specific implementations, the first indicators can also be divided into more or fewer categories. The present application places no restrictions on the types of the first indicators.

[0271] For the case where the first indicator includes multiple indicators, in the above embodiments (such as Figure 4 , Figure 6 , Figure 7 and other illustrated embodiments), the process of generating an evaluation environment according to the first indicator and performing an evaluation based on the evaluation environment can be understood as that multiple first indicators implement this process in parallel. In a possible design, for multiple first indicators, the above process of generating an evaluation environment according to the first indicator and performing an evaluation based on the evaluation environment can also be implemented in a serial manner. For example, S201 - S203 may include: sequentially for each first indicator, generating an evaluation environment corresponding to the first indicator according to the first indicator, the scenario information of the application scenario, and the entity information of the first entity in the application scenario; invoking a trust model to perform a trust evaluation on the first entity based on the evaluation environment corresponding to the first indicator, and outputting a trust evaluation result of the first entity; and outputting an evaluation result of the trust model according to the trust evaluation result corresponding to the first indicator.

[0272] In other words, in this design, the trust evaluation ability of the trust model can be evaluated for one first indicator first; after completing the trust model evaluation corresponding to one first indicator, the trust model evaluation corresponding to the next first indicator can be continued until the trust model evaluations corresponding to all first indicators are completed. That is, each first indicator corresponds to an evaluation environment, and the evaluation environments corresponding to different first indicators are different.

[0273] Exemplarily, taking the process of generating an evaluation environment according to the first indicator and performing an evaluation based on the evaluation environment in the above Figure 6 illustrated embodiment as an example (it can also be Figure 4 or Figure 7 and other illustrated embodiments), Figure 8 shows another implementation flowchart of the trust model evaluation method provided by the embodiments of the present application. As Figure 8 shown, this trust model evaluation method may include S801 - S816.

[0274] S801. The interface module receives the application scenario and the first entity input by the user.

[0275] S802. The interface module sends the scenario information of the application scenario and the entity information of the first entity to the analysis module.

[0276] Correspondingly, the analysis module receives the scenario information of the application scenario and the entity information of the first entity.

[0277] S803. The analysis module extracts the entity features of the first entity and the scenario features of the application scenario.

[0278] S804. The analysis module sends the entity features of the first entity and the scenario features of the application scenario to the fusion module.

[0279] Correspondingly, the fusion module receives the entity features of the first entity and the scenario features of the application scenario.

[0280] S805. The fusion module fuses the entity features of the first entity and the scenario features of the application scenario to obtain fused features.

[0281] S806. The fusion module sends the fused features to the metric generation module.

[0282] Correspondingly, the metric generation module receives the fused features.

[0283] S807. The metric generation module outputs a first metric that matches the application scenario and the first entity based on the fused features.

[0284] S808. The metric generation module sends the first metric that matches the application scenario and the first entity to the metric storage module.

[0285] Correspondingly, the metric storage module receives the first metric that matches the application scenario and the first entity.

[0286] S801 - S808 can be referred to the above S601 - S608 and will not be elaborated here.

[0287] S809. The fusion module obtains a first metric from the metric storage module.

[0288] It can be understood that the first metric is any metric that matches the application scenario and the first entity.

[0289] S810. The fusion module generates an evaluation environment corresponding to the first metric based on the first metric, the entity features of the first entity, and the scenario features of the application scenario.

[0290] S811. The fusion module sends information about the evaluation environment corresponding to the first metric to the evaluation module.

[0291] Correspondingly, the evaluation module receives information about the evaluation environment corresponding to the first metric.

[0292] S812. The evaluation module calls the trust model to perform a trust evaluation on the first entity based on the evaluation environment corresponding to the first metric.

[0293] S813. The trust model returns a trust evaluation result corresponding to the first metric to the evaluation module.

[0294] S814. The evaluation module sends the trust evaluation result corresponding to the first metric to the parsing module.

[0295] Correspondingly, the parsing module receives the trust evaluation result corresponding to the first metric.

[0296] S815. The parsing module outputs the evaluation result of the trust model corresponding to the first metric according to the trust evaluation result corresponding to the first metric.

[0297] S809 - S815 can refer to the above S609 - S615. The difference is that each step in S809 - S815 is implemented for one type of first metric and will not be elaborated here.

[0298] S816. For the next type of first metric, repeat S809 - S815 until all the first metrics are evaluated.

[0299] For example, the next metric can be obtained, and according to the process of S810 - S815, the evaluation result of the trust model corresponding to the next type of first metric can be output; and so on in a loop until the evaluation results of the trust models corresponding to all the first metrics are obtained.

[0300] In this design, for multiple types of first metrics, the process of generating an evaluation environment according to the first metric and performing an evaluation based on the evaluation environment is implemented in a serial manner. Focusing on a single metric (evaluating only one specific type of first metric at a time) can make the evaluation process more concise, clear, and easy to understand and explain. Additionally, implementing the process of generating an evaluation environment according to the first metric and performing an evaluation based on the evaluation environment in a serial manner for multiple types of first metrics can also enhance the evaluation depth and improve the evaluation accuracy. For example, resources and attention can be concentrated on one metric, avoiding the noise and interference that may be brought by multiple metrics, and more clearly understanding and analyzing the importance and role of each metric.

[0301] In a possible design, the evaluation environment generated in the above embodiments may include an initial evaluation environment and at least one other evaluation environment; the initial evaluation environment is generated according to the first metric, the scenario information, and the entity information of the first entity, and specific details can refer to the foregoing embodiments. The other evaluation environment can be obtained by adjusting the parameters in the initial evaluation environment. For example, the scenario features and entity features in the initial evaluation environment can be changed or adjusted to obtain the other evaluation environment.

[0302] Exemplarily, in this design, the method further includes: adjusting the parameters of the evaluation environment to obtain at least one updated evaluation environment; invoking the trust model, and performing a trust evaluation on the first entity based on the updated evaluation environment, and outputting the trust evaluation result of the first entity. Wherein, the evaluation environment is the initial evaluation environment, and the updated evaluation environment is the other evaluation environment.

[0303] In this design, the evaluation environment can be extended to provide a richer evaluation environment for the evaluation of the trust model, thereby improving the accuracy of the trust model evaluation. For example, the trust model can be evaluated based on multiple evaluation environments such as the initial evaluation environment and other evaluation environments, and the performance of the trust model can be comprehensively evaluated according to the evaluation results obtained from different evaluation environments.

[0304] In a possible design, the evaluation result of the trust model described in the above embodiments can also be used to indicate the update strategy of the trust model.

[0305] Exemplarily, as described in the foregoing embodiments, the evaluation result of the trust model can indicate the trust evaluation ability of the trust model within the dimension of the first metric. For different first metrics, the trust evaluation ability of the trust model may be relatively good within some first metric dimensions and relatively poor within some other first metric dimensions. The evaluation result of the trust model can indicate that it is necessary to optimize or update the trust model for the first metric with relatively poor trust evaluation ability to improve the performance of the trust model in such first metric dimensions. That is to say, the update strategy of the trust model can include which dimensions of the trust model need to be optimized or updated. For example, the update strategy can indicate that the security of the trust model is insufficient.

[0306] Optionally, the update strategy of the trust model can also indicate the specific way to update the trust model, or suggestions for improvement or optimization. For example, when it is necessary to optimize the robustness of the trust model, the update strategy can also indicate using more training data to train the trust model. Another example is that when it is necessary to optimize the security of the trust model, the update strategy can also indicate the specific way to improve the design of the trust evaluation curve, etc.

[0307] Exemplarily, the evaluation result of the trust model can include specific indication information, and this indication information can be used to indicate the update strategy of the trust model.

[0308] In this design, the evaluation result of the trust model can indicate the update strategy of the trust model, which can point out the improvement direction (dimension) of the trust model for developers or users, or provide more detailed model optimization suggestions for users, so that users can further optimize or adjust the trust model according to the evaluation result of the trust model to improve the performance of the trust model.

[0309] In other words, this design can provide feedback and guidance for the optimization of the trust model.

[0310] In addition, by updating the trust model according to the evaluation results of the trust model, the generalization performance of the trust model for different application scenarios can also be improved. For example, for the same trust model, after updating the trust model in the manner of this design in different application scenarios, the trust model can exhibit better performance in different application scenarios and has stronger application scenario adaptability.

[0311] Optionally, the trust model evaluation method provided in the embodiments of the present application may further include: updating the trust model according to the evaluation results of the trust model.

[0312] Exemplarily, the update method of the trust model may refer to the indication of the update strategy described in the foregoing embodiments, which will not be elaborated herein.

[0313] Taking the above Figure 4 illustrated embodiment as an example, Figure 9 Fig. shows another schematic flowchart of the implementation of the trust model evaluation method provided in the embodiments of the present application. As Figure 9 shown, the trust model evaluation method may include S901 - S911.

[0314] S901. The interface module receives the application scenario and the first entity input by the user.

[0315] S902. The interface module sends the scenario information of the application scenario and the entity information of the first entity to the analysis module.

[0316] Correspondingly, the analysis module receives the scenario information of the application scenario and the entity information of the first entity.

[0317] S903. The analysis module extracts the entity features of the first entity and the scenario features of the application scenario.

[0318] S904. The analysis module sends the entity features of the first entity and the scenario features of the application scenario to the fusion module.

[0319] Correspondingly, the fusion module receives the entity features of the first entity and the scenario features of the application scenario.

[0320] S905. The fusion module obtains the first metric from the metric storage module.

[0321] S906. The fusion module generates an evaluation environment according to the first metric, the entity features of the first entity, and the scenario features of the application scenario.

[0322] S907. The fusion module sends the information of the evaluation environment to the evaluation module.

[0323] Correspondingly, the evaluation module receives the information of the evaluation environment.

[0324] S908. The evaluation module invokes the trust model and conducts a trust evaluation on the first entity based on the evaluation environment.

[0325] S909. The trust model returns the trust evaluation result corresponding to the first metric to the evaluation module.

[0326] S910. The evaluation module sends the trust evaluation result corresponding to the first metric to the parsing module.

[0327] Correspondingly, the parsing module receives the trust evaluation result corresponding to the first metric.

[0328] S911. The parsing module outputs the evaluation result of the trust model according to the trust evaluation result corresponding to the first metric.

[0329] S901 - S911 can be referred to the above - mentioned S401 - S411 and will not be elaborated here.

[0330] S912. The parsing module sends the evaluation result of the trust model to the trust model.

[0331] Correspondingly, the trust model receives the evaluation result.

[0332] S913. The trust model is updated according to the evaluation result.

[0333] It can be understood that the trust model shown in the figure can be a software module for storing or managing the trust model, such as a model maintenance module, which can receive the evaluation result and update the trust model according to the evaluation result.

[0334] In this design, updating the trust model according to the evaluation result can improve the trust evaluation ability of the trust model.

[0335] In a possible design, the evaluation result of the trust model described in the above embodiments can be a semantic trust evaluation report.

[0336] Exemplarily, the trust evaluation result corresponding to the first metric can be input into a large - language model (LLM), and the LLM parses the trust evaluation result corresponding to the first metric to output a semantic trust evaluation report as the evaluation result of the trust model.

[0337] For example, the above - mentioned parsing module can be deployed on a semantic information processor, and the parsing module or the semantic information processor can use the LLM to process semantic - level information and parse the trust evaluation result corresponding to the first metric to generate a semantic trust evaluation report.

[0338] Optionally, the semantic trust evaluation report can provide improvement suggestions for the trust model, such as providing semantic adjustment solutions and optimization methods, to facilitate users to improve and enhance the trust model.

[0339] In this design, using the semantic trust evaluation report as the evaluation result of the trust model enables users to better understand the trust evaluation ability of the trust model within the dimension of the first indicator, and more intuitively understand from which dimensions the trust model can be improved, and / or, in what ways the trust model can be improved.

[0340] In some other embodiments, in the above-mentioned trust model evaluation method, the functions respectively implemented by the steps of extracting scenario features and entity features, and outputting the evaluation result of the trust model according to the trust evaluation result of the first entity, etc., can be integrated and implemented in one module. For example, such as an artificial intelligence (AI) module.

[0341] This application does not limit the division of the functional modules for implementing the trust model evaluation method.

[0342] It should be understood that the above mainly introduces the solution provided by the embodiments of this application from the perspective of the logic between the various steps of the trust model evaluation method. It can be understood that, in order to implement the functions of the above method, the embodiments of this application also provide a device for implementing the functions of the above trust model evaluation method, such as a trust model evaluation device. The functions of this device can be implemented by hardware or by hardware executing corresponding software. The hardware or software includes one or more units or modules corresponding to the functions of the above method, that is, this device can include the corresponding hardware structure and / or software module for executing each function of the above method. This device can be an electronic device, or a chip built in an electronic device or a functional module in an electronic device. The electronic device can refer to the electronic device described in the foregoing embodiments.

[0343] Exemplarily, Figure 10 shows a schematic composition diagram of the trust model evaluation device provided by the embodiments of this application. As Figure 10 shown, the trust model evaluation device can include: a trust evaluation unit 1001 and a model evaluation unit 1002.

[0344] Among them, the trust evaluation unit 1001 is used to perform a trust evaluation on the first entity based on the trust model to obtain the trust evaluation result of the first entity. The trust evaluation result includes at least one group, and each group of trust evaluation results corresponds to a first indicator, and the first indicator is used to indicate the evaluation dimension of the trust model.

[0345] The model evaluation unit 1002 evaluates the trust model according to the trust evaluation result to obtain the evaluation result of the trust model, and the evaluation result of the trust model is used to indicate the trust evaluation ability of the trust model within the dimension of the first metric.

[0346] Exemplarily, the trust evaluation unit 1001 may include the analysis module, the fusion module, the evaluation module, the metric generation / storage module, etc. described in the foregoing embodiments. The model evaluation unit 1002 may include the parsing module described in the foregoing embodiments.

[0347] In a possible design, the trust evaluation unit 1001 is specifically configured to: obtain the first metric, the scenario information of the application scenario, and the entity information of the first entity; generate an evaluation environment according to the first metric, the scenario information of the application scenario, and the entity information of the first entity, where the evaluation environment indicates the environment in which the first entity runs in the application scenario; call the trust model, and perform a trust evaluation on the first entity based on the evaluation environment to obtain the trust evaluation result of the trust model for the first entity.

[0348] In a possible design, the trust evaluation unit 1001 is specifically configured to: obtain the scenario information of the application scenario and the entity information of the first entity; and obtain the first metric based on the scenario information of the application scenario and the entity information of the first entity.

[0349] In a possible design, the first metric used to generate the evaluation environment may be predefined, or preconfigured, or configured. For example, evaluation metrics matching the application scenario and the entity may be configured to obtain the mapping relationship between the evaluation metrics, the application scenario, and the entity. The trust evaluation unit 1001 is specifically configured to: determine the evaluation metrics matching the application scenario and the first entity according to this mapping relationship, such as the so-called target evaluation metrics, and select at least one metric from the target evaluation metrics as the first metric.

[0350] In another possible design, the first metric used to generate the evaluation environment may also be dynamically determined according to the scenario information and the entity information of the first entity. For example, the trust evaluation unit 1001 is further configured to dynamically determine the evaluation metrics matching the application scenario and the first entity, such as the so-called target evaluation metrics, by means of deep learning or machine learning, and select at least one metric from the target evaluation metrics as the first metric by using a pre-trained neural network model according to the scenario information and the entity information of the first entity.

[0351] In another possible design, the first metric used to generate the evaluation environment can also be one or more first metrics selected by the user from the obtained first metrics based on the scenario information of the application scenario and the entity information of the first entity. For example, the trust evaluation unit 1001 is further configured to, in response to the user's selection operation on the first metric, use the first metric indicated by the selection operation as the first metric used to generate the evaluation environment.

[0352] In a possible design, the trust evaluation unit 1001 is specifically configured to: fuse the scenario information of the application scenario and the entity information of the first entity to obtain a fused feature; and obtain the first metric based on the fused feature.

[0353] In a possible design, the trust evaluation unit 1001 is specifically configured to, for each of the first metrics in sequence, generate an evaluation environment corresponding to the first metric according to the first metric, the scenario information of the application scenario, and the entity information of the first entity in the application scenario; perform a trust evaluation on the first entity based on the evaluation environment corresponding to the first metric, and output a trust evaluation result of the first entity.

[0354] That is, in this design, each first metric corresponds to an evaluation environment, and the evaluation environments corresponding to different first metrics are different.

[0355] In a possible design, the generated evaluation environment may include an initial evaluation environment and at least one other evaluation environment; the initial evaluation environment is generated according to the first metric, the scenario information, and the entity information of the first entity; the other evaluation environment may be obtained by adjusting the parameters in the initial evaluation environment. For example, the trust evaluation unit 1001 is further configured to adjust the parameters of the evaluation environment to obtain at least one updated evaluation environment; call the trust model, perform a trust evaluation on the first entity based on the updated evaluation environment, and output a trust evaluation result of the first entity. Here, the evaluation environment is the initial evaluation environment, and the updated evaluation environment is the other evaluation environment.

[0356] In a possible design, the application scenario includes at least one of the following: network security scenario, intelligent system scenario, online platform scenario, complex decision-making environment scenario, Internet of Things scenario, cloud computing scenario, edge computing scenario, network slicing and virtualized network scenario, software-defined network scenario, blockchain scenario, communication network scenario; the first entity includes at least one of the following: a hardware device or module, an application or software module, and data of hardware and / or software interaction.

[0357] In a possible design, the evaluation result of the trust model is also used to indicate the update strategy of the trust model.

[0358] In a possible design, the device further includes: an update unit (Figure 10 not shown in the figure) for updating the trust model according to the evaluation result of the trust model.

[0359] In a possible design, the evaluation result of the trust model is a semantic trust evaluation report.

[0360] In a possible design, the first indicator includes one or more of the following indicators: data security, comprehensiveness of the trust evaluation result, usability of the trust model, functionality of the trust model, robustness of the trust model, neutrality of the trust evaluation result, and interpretability of the trust evaluation result.

[0361] It should be understood that the division of units in the above device is only a logical function division. In actual implementation, it can be fully or partially integrated into a physical entity, or physically separated. And the units in the device can all be implemented in the form of software called by processing elements; they can also all be implemented in the form of hardware; or some units can be implemented in the form of software called by processing elements, and some units can be implemented in the form of hardware.

[0362] For example, each unit can be a separately established processing element, or can be integrated in a certain chip of the device. In addition, it can also be stored in the memory in the form of a program and called and executed by a certain processing element of the device to perform the function of the unit. In addition, these units can be fully or partially integrated together, or can be independently implemented. The processing element mentioned here can also be called a processor, which can be an integrated circuit with signal processing capabilities. In the implementation process, each step of the above method or each of the above units can be implemented through the integrated logic circuit in the processor element or in the form of software called by the processing element.

[0363] In an example, the units in any of the above devices can be one or more integrated circuits configured to implement the above method, for example: one or more application specific integrated circuits (ASICs), or, one or more digital signal processing (DSP) circuits, or, one or more field programmable gate arrays (FPGAs), or a combination of at least two of these integrated circuit forms.

[0364] Again, when the units in the device can be implemented in the form of a processing element scheduler, the processing element can be a general-purpose processor, such as a CPU or other processor that can call a program. Again, these units can be integrated together and implemented in the form of a system-on-a-chip (SOC).

[0365] When the device includes a receiving unit, the receiving unit in the device is an interface circuit or an input circuit of the device for receiving signals from other devices. For example, when the device is implemented in the form of a chip, the receiving unit is an interface circuit or an input circuit of the chip for receiving signals from other chips or devices. When the device includes a transmitting unit, the transmitting unit is an interface circuit or an output circuit of the device for transmitting signals to other devices. For example, when the device is implemented in the form of a chip, the transmitting unit is an interface circuit or an output circuit of the chip for transmitting signals to other chips or devices.

[0366] For example, an embodiment of the present application may further provide a trust model evaluation device, which can be applied to the above-mentioned electronic device. The trust model evaluation device may include: a processor and an interface circuit. The processor may include one or more. The processor is used to communicate with other devices through the interface circuit and execute each step performed by the electronic device in the above method.

[0367] In one implementation, the units of the electronic device that implement each corresponding step in the above method may be implemented in the form of a processing element scheduler. For example, the device for the electronic device may include a processing element and a storage element. The processing element calls the program stored in the storage element to execute the method performed by the electronic device in the above method embodiment. The storage element may be a storage element on the same chip as the processing element, that is, an on-chip storage element.

[0368] In another implementation, the program for executing the above method may be stored in a storage element on a different chip from the processing element, that is, an off-chip storage element. At this time, the processing element calls or loads the program from the off-chip storage element onto the on-chip storage element to call and execute the method in the above method embodiment.

[0369] For example, an embodiment of the present application may further provide a trust model evaluation device, which may include a processor for executing computer instructions stored in a memory. When the computer instructions are executed, the device executes the method performed by the above-mentioned electronic device. The memory may be located inside or outside the trust model evaluation device. And the processor includes one or more.

[0370] In yet another implementation, the units of each step in the above method may be configured as one or more processing elements, and these processing elements may be correspondingly arranged on the electronic device. Here, the processing element may be an integrated circuit, for example: one or more ASICs, or one or more DSPs, or one or more FPGAs, or a combination of these types of integrated circuits. These integrated circuits may be integrated together to form a chip.

[0371] The units for implementing each step in the above method can be integrated together and implemented in the form of a SOC. This SOC chip is used to implement the corresponding method. At least one processing element and a storage element can be integrated in this chip, and the corresponding method is implemented in the form of a program stored in the storage element being called by the processing element; alternatively, at least one integrated circuit can be integrated in this chip for implementing the corresponding method; alternatively, the above implementation manners can be combined, and the functions of some units are implemented in the form of a program called by the processing element, and the functions of some units are implemented in the form of an integrated circuit.

[0372] The processing element here is the same as the above description. It can be a general-purpose processor, such as a CPU, or can also be one or more integrated circuits configured to implement the above method, such as: one or more ASICs, or, one or more microprocessors DSPs, or, one or more FPGAs, etc., or a combination of at least two of these integrated circuit forms.

[0373] The storage element can be a memory or a collective term for multiple storage elements.

[0374] For example, the embodiments of the present application also provide a chip system, and this chip system can be applied to the above-mentioned electronic device. The chip system includes one or more interface circuits and one or more processors; the interface circuits and the processors are interconnected by lines; the processors receive and execute computer instructions from the memory of the electronic device through the interface circuits to implement the method in the above method embodiments.

[0375] Through the description of the above implementation manners, those skilled in the art can clearly understand that, for the convenience and simplicity of description, only the above division of each functional module is used as an example for illustration. In actual applications, the above functions can be allocated to different functional modules according to needs, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above.

[0376] In several embodiments provided by the present application, it should be understood that the disclosed device and method can be implemented in other ways. For example, the device embodiments described above are only illustrative. For example, the division of the modules or units is only 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 device, or some features can be ignored or not executed. Another point, 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 the device or unit can be in an electrical, mechanical or other form.

[0377] The unit described as a separation component may or may not be physically separated. The component shown as a unit may be a single physical unit or multiple physical units, that is, it may be located in one place or distributed to multiple different places. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0378] In addition, each functional unit in various embodiments of the present application may be integrated in a processing unit, may exist separately as individual physical units, or two or more units may be integrated in one unit. The above integrated units can be implemented in the form of hardware or in the form of software functional units.

[0379] If the 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 readable storage medium. Based on such an understanding, the technical solution of the embodiments of the present application, in essence, or the part that contributes to the prior art, or all or part of this technical solution, can be embodied in the form of a software product, such as a program. This software product is stored in a program product, such as a computer-readable storage medium, and includes several instructions to enable a device (which can be a single-chip microcomputer, a chip, etc.) or a processor to execute all or part of the steps of the methods described in various embodiments of the present application. The aforementioned storage medium includes: USB flash drives, mobile hard disks, ROM, RAM, magnetic disks, or optical discs and other various media that can store program codes.

[0380] For example, the embodiments of the present application can also provide a computer-readable storage medium, including: computer software instructions; when the computer software instructions run on an electronic device or in a chip built into the electronic device, the electronic device can be enabled to execute the method as described in the foregoing embodiments.

[0381] Optionally, the embodiments of the present application also provide a trust model evaluation device, including: a transceiver unit and a processing unit. The transceiver unit can be used to send and receive information, or to communicate with other network elements (such as terminal devices or network devices, or other electronic devices or apparatuses). The processing unit can be used to process data. This device can implement the method as described in the above embodiments through the transceiver unit and the processing unit.

[0382] Optionally, the embodiments of the present application also provide a computer program product, which can implement the method as described in the above embodiments when executed.

[0383] Based on the above embodiments, the embodiments of the present application also provide an artificial intelligence model, and the artificial intelligence model has the function of implementing the method as described in the above embodiments.

[0384] Optionally, the artificial intelligence model includes one or more models. When the artificial intelligence model includes multiple models, the multiple models implement different functions in the method described in the foregoing embodiments.

[0385] An embodiment of the present application further provides a communication system, including: a first entity and a second entity, where the second entity interacts with the first entity to implement the method described in the foregoing embodiments.

[0386] As described above, the foregoing are only specific embodiments of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions within the technical scope disclosed in the present application should be covered by the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for evaluating a trust model, characterized in that, The method includes: Performing a trust assessment on a first entity based on a trust model to obtain a trust assessment result of the first entity, where the trust assessment result includes at least one group, and each group of the trust assessment results corresponds to a first metric, and the first metric is used to indicate an evaluation dimension of the trust model; Evaluating the trust model according to the trust assessment result to obtain an evaluation result of the trust model, where the evaluation result of the trust model is used to indicate the trust assessment ability of the trust model within the dimension of the first metric.

2. The method according to claim 1, characterized in that, The performing a trust assessment on a first entity based on a trust model to obtain a trust assessment result of the first entity includes: Obtaining the first metric, scenario information of an application scenario, and entity information of the first entity; Generating an evaluation environment according to the first metric, the scenario information of the application scenario, and the entity information of the first entity, where the evaluation environment indicates an environment in which the first entity runs in the application scenario; Invoking the trust model to perform a trust assessment on the first entity based on the evaluation environment to obtain a trust assessment result of the trust model for the first entity.

3. The method according to claim 2, characterized in that, The obtaining the first metric, scenario information of an application scenario, and entity information of the first entity includes: Obtaining the scenario information of the application scenario and the entity information of the first entity; Obtaining the first metric based on the scenario information of the application scenario and the entity information of the first entity.

4. The method according to claim 3, characterized in that, The obtaining the first metric based on the scenario information of the application scenario and the entity information of the first entity includes: Fusing the scenario information of the application scenario and the entity information of the first entity to obtain a fused feature; Obtaining the first metric based on the fused feature.

5. The method according to any one of claims 2 - 4, characterized in that, The application scenario includes at least one of the following: network security scenario, intelligent system scenario, online platform scenario, complex decision-making environment scenario, Internet of Things scenario, cloud computing scenario, edge computing scenario, network slicing and virtualized network scenario, software-defined network scenario, blockchain scenario, communication network scenario; The first entity includes at least one of the following: a hardware device or module, an application or software module, and data of hardware and / or software interaction.

6. The method according to any one of claims 2 - 5, characterized in that, The entity information of the first entity includes static attribute information of the first entity and / or dynamic behavior information.

7. The method according to any one of claims 1 - 6, characterized in that, The evaluation result of the trust model is also used to indicate an update strategy of the trust model.

8. The method according to any one of claims 1 - 7, characterized in that, The method further includes: Updating the trust model according to the evaluation result of the trust model.

9. The method according to any one of claims 1 - 8, characterized in that, The evaluation result of the trust model is a semantic trust assessment report.

10. The method according to any one of claims 1 - 9, characterized in that, The first metric includes one or more of the following metrics: security of data, comprehensiveness of trust assessment result, usability of trust model, functionality of trust model, robustness of trust model, neutrality of trust assessment result, and interpretability of trust assessment result.

11. A trust model evaluation device, characterized in that, The apparatus includes a module for executing the method according to any one of claims 1-10.

12. A trust model evaluation device, characterized in that, The apparatus includes: a processor configured to execute the method according to any one of claims 1-10.

13. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes instructions that, when executed, cause the method according to any one of claims 1-10 to be implemented.

14. A computer program product, characterized in that, When the computer program product is executed, the method according to any one of claims 1-10 is caused to be implemented.

15. A chip system, characterized in that, The chip system includes one or more interface circuits and one or more processors; the interface circuits and the processors are interconnected by lines; the processors receive and execute computer instructions from the memory of the electronic device through the interface circuits to implement the method according to any one of claims 1-10.

Images