Communication interaction method and device based on trusted token mechanism, equipment and medium

Abstract

Embodiments of the present disclosure disclose a communication interaction method and device based on a trusted token mechanism, an apparatus, and a medium. A specific implementation of the method includes: in response to receiving interaction information, querying whether there is trusted token mechanism association information; in response to the absence of corresponding trusted token mechanism association information, generating simulated trusted token information for each trusted token mechanism determined; filtering out a primary trusted token mechanism and a set of alternative trusted token mechanisms; performing communication interaction processing; and in response to obtaining actual communication interaction records within a target time period, storing the determined trusted token mechanism association information and the interaction scenario to a trusted token mechanism association repository. This implementation uses a multi-modal large model to accurately and efficiently filter out an initial trusted token mechanism that is suitable for an interaction scenario. On this basis, through simulated operation and real interaction record analysis, further determination of the corresponding trusted token mechanism in the interaction scenario is achieved.

Description

Technical Field

[0001] The embodiments of this disclosure relate to the field of computer technology, and more specifically to communication interaction methods, apparatus, devices, and media based on a trust token mechanism. Background Technology

[0002] Currently, the trust token mechanism is a core mechanism widely used in identity verification, authorization, and resource access control. Its core principle is to achieve secure communication through encrypted temporary credentials. In real-world scenarios, different scenarios or interfacing organizations often employ different trust token mechanisms. The selection and storage of trust token mechanisms typically involves manual selection and storage by relevant technical personnel based on historical experience.

[0003] However, when using the above method, the following technical problems often arise: Given the complexity of communication interaction scenarios, different scenarios exhibit different communication characteristics. Manually selecting a trust token mechanism is not only inefficient, but an unsuitable mechanism can also lead to communication security or efficiency issues during the interaction process. Therefore, effective management of trust token mechanisms is currently a major research direction.

[0004] The information disclosed in this background section is only intended to enhance the understanding of the background of the inventive concept, and therefore may contain information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0005] The summary portion of this disclosure is intended to provide a brief overview of the concepts, which will be described in detail in the detailed description portion. This summary portion is not intended to identify key or essential features of the claimed technical solutions, nor is it intended to limit the scope of the claimed technical solutions.

[0006] Some embodiments of this disclosure propose communication interaction methods, apparatuses, devices, and media based on a trust token mechanism to solve one or more of the technical problems mentioned in the background section above.

[0007] In a first aspect, some embodiments of this disclosure provide a communication interaction method based on a trust token mechanism, comprising: in response to receiving interaction information for communication interaction with a target external organization, determining an interaction scenario corresponding to the interaction information; querying a trust token mechanism association repository to determine whether there is trust token mechanism association information corresponding to the interaction scenario, wherein the trust token mechanism association information includes: a trust token mechanism, a trust token generation basis, a trust token storage method, and security priority information; in response to determining that there is no corresponding trust token mechanism association information, generating at least one trust token mechanism for the interaction scenario based on various interaction event information under the interaction scenario; and for each trust token mechanism, based on the aforementioned trust order... The system simulates the use of a trust token mechanism in the above-mentioned interactive scenarios, generating simulated credit indicator information. Based on at least one simulated credit indicator information, a target number of trust token mechanisms are selected from the at least one trust token mechanism, serving as a primary trust token mechanism and a set of alternative trust token mechanisms. Based on the primary trust token mechanism and the set of alternative trust token mechanisms, communication interaction processing in the above-mentioned interactive scenarios is executed. In response to obtaining the actual communication interaction records within the target time period, the system determines the associated information of the trust token mechanism matched in the above-mentioned interactive scenarios, and stores the associated information of the matched trust token mechanism and the above-mentioned interactive scenarios in the form of key-value pairs in the above-mentioned trust token mechanism associated repository.

[0008] Secondly, some embodiments of this disclosure provide a communication interaction device based on a trust token mechanism, including: a first determining unit configured to determine an interaction scenario corresponding to the interaction information in response to receiving interaction information for communication interaction with a target external organization; a querying unit configured to query a trust token mechanism association repository to determine whether there is trust token mechanism association information corresponding to the interaction scenario, wherein the trust token mechanism association information includes: trust token mechanism, trust token generation basis, trust token storage method, and security priority information; a first generating unit configured to generate at least one trust token mechanism in the interaction scenario based on various interaction event information in the interaction scenario in response to determining that there is no corresponding trust token mechanism association information; and a second generating unit configured to, for each trust token mechanism, based on the... The above-described simulation method for the credit token mechanism generates simulated credit indicator information simulating the use of the credit token mechanism in the above-described interaction scenario; a filtering unit is configured to filter a target number of credit token mechanisms from the above-described at least one credit token mechanism based on at least one simulated credit indicator information, respectively serving as a primary credit token mechanism and a set of alternative credit token mechanisms; an execution unit is configured to execute communication interaction processing in the above-described interaction scenario based on the primary credit token mechanism and the set of alternative credit token mechanisms; a second determining unit is configured to, in response to the actual communication interaction records obtained within the target time period, determine the credit token mechanism association information matched in the above-described interaction scenario, and store the matched credit token mechanism association information and the above-described interaction scenario in the form of key-value pairs in the above-described credit token mechanism association repository.

[0009] Thirdly, some embodiments of this disclosure provide an electronic device, including: one or more processors; and a storage device having one or more programs stored thereon, such that when the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in any implementation of the first aspect.

[0010] Fourthly, some embodiments of this disclosure provide a computer-readable medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the method as described in any implementation of the first aspect.

[0011] The above embodiments of this disclosure have the following beneficial effects: Through the communication interaction method based on a trust token mechanism in some embodiments of this disclosure, a multimodal large model can be used to accurately and efficiently filter out the initial trust token mechanism suitable for the interaction scenario. Based on this, through simulated operation and analysis of real interaction records, the corresponding trust token mechanism for the interaction scenario is further determined. Specifically, the reasons for problems or low efficiency in related communication interactions are: due to the complexity of communication interaction scenarios, different communication interaction scenarios have different communication interaction characteristics. Manually selecting a trust token mechanism is not only inefficient, but the selection of an unsuitable trust token mechanism can lead to communication security or efficiency problems during the communication interaction process. Furthermore, effective control of the trust token mechanism is one of the main research directions at present. Based on this, the communication interaction method based on a trust token mechanism in some embodiments of this disclosure first, in response to receiving interaction information for communication interaction with a target external organization, determines the interaction scenario corresponding to the aforementioned interaction information. Here, by determining the interaction scenario, suitable token mechanism association information for communication interaction with the target external organization can be selected according to the interaction characteristics of the interaction scenario. Then, the system queries the trust token mechanism association repository to determine if there is any trust token mechanism association information corresponding to the aforementioned interaction scenario. This association information includes: the trust token mechanism, the basis for trust token generation, the trust token storage method, and security priority information. By setting up the trust token mechanism association repository, the system can store the association information for various applicable trust token mechanisms under different interaction scenarios, effectively helping to determine the trust token mechanism for each scenario. Furthermore, by setting the trust token mechanism association information to include the trust token mechanism, the basis for trust token generation, the trust token storage method, and security priority information, the system can select the trust token mechanism from multiple perspectives, choosing the most suitable trust token mechanism for the interaction scenario based on matching different characteristics. Next, in response to the determination that no corresponding trust token mechanism association information exists, at least one trust token mechanism for the aforementioned interaction scenario can be accurately generated based on the interaction event information. Here, even when no trust token mechanism association information is found, by fully understanding the interaction characteristics and scenario details through the content of each event in the interaction scenario, at least one trust token mechanism can be accurately preliminarily selected. Next, for each credit token mechanism, based on the simulation method corresponding to the aforementioned credit token mechanism, simulated credit indicator information simulating the use of the credit token mechanism in the aforementioned interaction scenario can be accurately generated, so as to determine the effectiveness and matching of each credit token mechanism through simulated credit indicator information.Furthermore, based on at least one simulated credit granting indicator, a target number of credit token mechanisms with the best matching features are selected from the at least one credit token mechanism mentioned above. These are designated as the primary credit token mechanism and the set of alternative credit token mechanisms, respectively, for subsequent use. Next, based on the primary and alternative credit token mechanisms, communication interaction processing under the aforementioned interaction scenario is executed. Finally, in response to obtaining the actual communication interaction records within the target time period, the associated information of the matched credit token mechanisms under the aforementioned interaction scenario is determined, and the associated information of the matched credit token mechanisms and the aforementioned interaction scenario are stored in the aforementioned credit token mechanism association repository in the form of key-value pairs. This enables the management of the credit token mechanism association information and the selection of credit token mechanisms for subsequent interaction scenarios. In summary, by utilizing a multimodal large model, the appropriate initial trust token mechanism for the interaction scenario can be accurately and efficiently selected. Based on this, through simulation and analysis of real interaction records, the corresponding trust token mechanism for the interaction scenario can be further determined, which can effectively avoid the security and efficiency problems in communication interaction caused by manual selection. Attached Figure Description

[0012] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and elements are not necessarily drawn to scale.

[0013] Figure 1 This is a flowchart of some embodiments of the communication interaction method based on the trust token mechanism according to this disclosure; Figure 2 This is a schematic diagram of the structure of some embodiments of a communication interaction device based on a trust token mechanism according to the present disclosure; Figure 3 This is a schematic diagram of the structure of an electronic device suitable for implementing some embodiments of the present disclosure. Detailed Implementation

[0014] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0015] It should also be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings. Unless otherwise specified, the embodiments and features described in this disclosure can be combined with each other.

[0016] It should be noted that the concepts of "first" and "second" mentioned in this disclosure are used only to distinguish different devices, modules or units, and are not used to limit the order of functions performed by these devices, modules or units or their interdependencies.

[0017] It should be noted that the terms "a" and "a plurality of" used in this disclosure are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0018] The names of messages or information exchanged between multiple devices in the embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.

[0019] This disclosure will now be described in detail with reference to the accompanying drawings and embodiments.

[0020] refer to Figure 1 The diagram illustrates a flow 100 of some embodiments of a communication interaction method based on a trust token mechanism according to the present disclosure. This communication interaction method based on a trust token mechanism includes the following steps: Step 101: In response to receiving interaction information for communication with the target external organization, determine the interaction scenario corresponding to the aforementioned interaction information.

[0021] In some embodiments, in response to receiving interaction information for communication with a target external institution, the executing entity (e.g., an electronic device) of the aforementioned communication interaction method based on a trust token mechanism can determine the interaction scenario corresponding to the interaction information. The target external institution can be an interaction object for communication. For example, the target external institution can be an institution in the financial sector or an institution in the e-commerce sector. Communication interaction can be the exchange of communication information. For example, communication interaction can be communication interaction under user authentication and authorization, inter-service communication, or communication interaction between a mobile terminal and an IoT device. Interaction information can be an interaction request for communication with the target external institution. The interaction request can be a request in a predetermined request format. The interaction scenario can be a scenario in which the executing entity interacts with the target external institution. For example, the interaction scenario can be one of the following: inter-service communication scenario, user authentication and authorization scenario, mobile terminal and IoT device interaction scenario, cross-domain and distributed system scenario.

[0022] As an example, the aforementioned executing entity can parse the request corresponding to the interaction information to obtain the interaction scenario.

[0023] Step 102: Query the trust token mechanism associated repository to see if there is any trust token mechanism associated information corresponding to the above-mentioned interaction scenario.

[0024] In some embodiments, the aforementioned executing entity can query a stored trust token mechanism association repository to see if there is trust token mechanism association information corresponding to the aforementioned interaction scenario. The trust token mechanism association information includes: trust token mechanism, trust token generation basis, trust token storage method, security priority information, and interaction scenario. The trust token mechanism association repository can be a database storing trust token mechanism association information. The trust token mechanism association information can include information including the trust token mechanism and other related information. The various trust token mechanisms stored in the trust token mechanism association repository are used for the selection and management of trust token mechanisms in various interaction scenarios, facilitating efficient selection of trust token mechanisms in the future. The trust token mechanism can be a mechanism for how to generate and transmit trust tokens to achieve authentication, authorization, and resource access control. The trust token mechanism can be, but is not limited to, one of the following: OAuth2 authorization code mode, OAuth2 user trust mode, SSO (JWT) mechanism, IoT authentication mechanism, and blockchain credit token mechanism. The trust token generation basis can be the basis for generating the trust token. For example, the generation basis for a trust token can be one of the following: for the OAuth2 authorization code mode, a user authorization code + client credentials; for the OAuth2 user trust mode, a user password + client credentials; for the SSO (JWT) mechanism, a user identity + permission statement; for the IoT authentication mechanism, a device fingerprint + pre-shared key; or for the blockchain trust token mechanism, smart contract rules + on-chain data. The trust token can be stored in any way. In practice, the trust token storage method can be, but is not limited to, one of the following: server-side session, client-side short-term caching, secure device storage, or decentralized wallet storage. The key security information can be the main security features of the trust token mechanism. For example, for the OAuth2 authorization code mode, the key security information is preventing authorization code leakage. For the OAuth2 user trust mode, the key security information is preventing password theft. For the SSO (JWT) mechanism, the key security information is preventing token tampering and hijacking. For the IoT authentication mechanism, the key security information is anti-cloning attacks and dynamic binding. For the blockchain trust token mechanism, the key security information is preventing replay attacks and compliance. In practice, the OAuth2 authorization code model corresponds to third-party login and API gateway scenarios. The OAuth2 user authorization model corresponds to internal management system scenarios. The SSO (JWT) mechanism corresponds to cross-system single sign-on scenarios. The IoT authentication mechanism corresponds to smart home and industrial sensor scenarios.

[0025] As an example, the aforementioned executing entity can generate a query statement corresponding to the interaction scenario to query whether there is any trust token mechanism associated information corresponding to the aforementioned interaction scenario in the trust token mechanism associated repository.

[0026] Step 103: In response to determining that there is no corresponding trust token mechanism associated information, at least one trust token mechanism for the above-mentioned interaction scenario is generated using a multimodal large model based on the interaction event information of each interaction scenario described above.

[0027] In some embodiments, in response to determining that no corresponding trust token mechanism association information exists, the executing entity can generate at least one trust token mechanism for the aforementioned interaction scenario based on the interaction event information of each interaction event in the aforementioned interaction scenario. The interaction event information can be the event content of interaction events that may occur in the interaction scenario. For example, for an interaction scenario of smart home, the corresponding interaction events could be: an event of a smart robot vacuum cleaning a target area, or an event of a smart air conditioner cooling. The trust token mechanism in at least one trust token mechanism can be a preliminarily determined trust token mechanism suitable for the interaction scenario. The multimodal large model can be a large model that supports multimodal content input and input. In practice, the multimodal large model is a large model trained on a training dataset corresponding to the trust-related task set. For example, the trust-related task can be a trust token mechanism generation task, a trust token mechanism evaluation task, or a trust token effect simulation task. The training dataset can include: a trust token mechanism generation dataset, a trust token mechanism evaluation dataset, and a trust token effect simulation dataset. The multimodal large model can be a large model based on the Transformer architecture or a large model based on the MoE architecture. The multimodal large model takes task processing prompts for credit-related tasks as input and outputs the task processing results for those tasks.

[0028] As an example, firstly, the aforementioned execution entity can generate a first generation prompt message to determine at least one trust token mechanism under the interaction scenario based on various interaction event information and interaction scenarios. Then, the first generation prompt message is input into the aforementioned multimodal large model to obtain at least one trust token mechanism.

[0029] In some optional implementations of certain embodiments, the execution entity may generate at least one authorization token mechanism for the aforementioned interaction scenario based on various interaction event information, including the following steps: The first step is to generate an interaction knowledge graph with multiple layers of nested content based on the aforementioned interaction event information. This interaction knowledge graph includes multiple interaction event chains, each with multiple connected branches. Each branch represents supplementary content during the interaction process. The interaction event chains can be in the form of chains describing the information of each interaction event. The multiple interaction event chains in the interaction knowledge graph support intersecting nodes. Here, the nodes in the interaction knowledge graph represent important event information. Edges represent the relationships between important event information. Using the interaction event chains as the main body, corresponding association information for each interaction event is added as supplementary edges to the interaction event chains, thereby constructing the interaction knowledge graph.

[0030] The second step involves retrieving a set of trust-granting event information and a set of classic scenario case information related to the aforementioned interaction scenarios from an external knowledge base. The external knowledge base can be an externally stored database of scenario knowledge related to various interaction scenarios. The trust-granting event information can be the event content of events that are related to the interaction scenario and involve trust-granting processing based on trust tokens. The classic scenario case information can be classic communication interaction cases based on trust tokens that occurred within the interaction scenario.

[0031] The third step involves generating at least one trust token mechanism for the aforementioned interaction scenario based on the aforementioned interaction knowledge graph, the aforementioned interaction event information, the aforementioned trust event information set, and the aforementioned scenario classic case information set, using a multimodal large model.

[0032] As an example, firstly, the aforementioned executing entity can generate a second generation prompt message based on the aforementioned interaction knowledge graph, the aforementioned information on each interaction event, the aforementioned set of trust event information, and the aforementioned set of classic case information for the scenario, to generate a trust token mechanism matching the interaction scenario. Then, the second generation prompt message is input into the multimodal large model to obtain at least one trust token mechanism under the aforementioned interaction scenario.

[0033] In some optional implementations of certain embodiments, the execution entity may generate at least one trust token mechanism for the aforementioned interaction scenario based on the aforementioned interaction knowledge graph, the aforementioned information on various interaction events, the aforementioned set of trust event information, and the aforementioned set of classic scenario case information, using a multimodal large model, including the following steps: The first step is to generate prompt information for the trust token mechanism based on the aforementioned interaction knowledge graph, the aforementioned interaction event information, the aforementioned trust event information set, and the aforementioned classic scenario case information set. Specifically, the prompt information for the trust token mechanism can be a prompt word generated by the multimodal large model based on the interaction knowledge graph, the aforementioned interaction event information, the aforementioned trust event information set, and the aforementioned classic scenario case information set.

[0034] The second step involves inputting the aforementioned trust token mechanism's generated prompt information into the aforementioned multimodal large model to obtain at least one first initial trust token mechanism. This first initial trust token mechanism can be a trust token mechanism initially generated by the multimodal large model, awaiting further verification to determine its accuracy.

[0035] Third, for at least one first initial trust token mechanism, perform the following first generation step: Sub-step 1 involves generating verification prompts for the feasibility, security, and matching of the at least one first initial trust token mechanism. These verification prompts may verify whether the first initial trust token mechanism is feasible and secure within the interaction scenario, and whether it matches the interaction scenario. The verification prompts may also include prompts for subsequent large-scale model verification of the feasibility, security, and matching of the first initial trust token mechanism.

[0036] Sub-step 2 involves inputting the aforementioned verification prompt information into a large model used for trust token mechanism vulnerability querying to obtain at least one trust token mechanism vulnerability information. The large model used for trust token mechanism vulnerability querying can be a large language model used for vulnerability querying of the initial trust token mechanism. Trust token mechanism vulnerability information can be vulnerabilities existing in the use of the trust token mechanism in interactive scenarios, such as security vulnerabilities or communication transmission vulnerabilities. In practice, the large model used for trust token mechanism vulnerability querying can be a large language model trained on a trust token mechanism training dataset, learning vulnerability-related knowledge corresponding to the trust token mechanism. The trust token mechanism training data can include the trust token mechanism and its corresponding vulnerability content. The large language model used for trust token mechanism vulnerability querying can be a large model based on the Transformer architecture.

[0037] Sub-step 3: In response to determining that at least one trust token mechanism vulnerability information corresponds to a vulnerability type that is not in a predetermined vulnerability type set and whose vulnerability level is lower than the target level, at least one first initial trust token mechanism is determined as at least one trust token mechanism. The vulnerability type can be a category of vulnerabilities existing in the trust token mechanism under the interaction scenario. Each vulnerability type in the predetermined vulnerability type set can be a pre-set type representing a serious vulnerability problem. The vulnerability level can represent the severity of the vulnerability existing under the interaction scenario. The higher the vulnerability level, the higher the severity of the vulnerability problem. The target level can be a pre-set vulnerability level. That is, vulnerability levels below the target level represent a lower vulnerability severity, while vulnerability levels above the target level represent a higher vulnerability severity. In practice, the predetermined vulnerability type can be, but is not limited to, one of the following: security vulnerability type, efficiency vulnerability type, or scenario adaptation vulnerability type.

[0038] Optionally, the steps also include: The first step involves, in response to determining that at least one trust token mechanism vulnerability information contains a target trust token mechanism vulnerability information whose corresponding vulnerability type falls within a predetermined vulnerability type set and / or whose vulnerability level is not lower than the aforementioned target level, adding at least one target trust token mechanism vulnerability information and its corresponding at least one first initial trust token mechanism to the aforementioned trust token mechanism generation prompt information, thereby obtaining addition prompt information. Here, adding at least one target trust token mechanism vulnerability information to the trust token mechanism generation prompt information provides a channel for subsequent multimodal large-scale models to re-establish the trust token mechanism based on the vulnerability content.

[0039] The second step involves inputting the aforementioned added prompt information into the aforementioned multimodal large model to obtain at least one second initial trust token mechanism. This second initial trust token mechanism can be a trust token mechanism regenerated from the multimodal large language model and combined with vulnerability information, awaiting further verification.

[0040] The third step is to use the above-mentioned at least one second initial trust token mechanism as at least one first initial trust token mechanism and continue to execute the above-mentioned generation steps.

[0041] Step 104: For each credit token mechanism, based on the simulation method corresponding to the credit token mechanism, generate simulated credit indicator information simulating the use of the credit token mechanism in the above-mentioned interaction scenario.

[0042] In some embodiments, for each of the at least one credit token mechanism described above, the executing entity can generate simulated credit indicator information simulating the use of the credit token mechanism in the above-mentioned interaction scenario based on the simulation method corresponding to the credit token mechanism. Different credit token mechanisms have corresponding simulation processing methods for simulating credit token processing. For example, for the OAuth2.0 authorization code mode, the corresponding simulation method is: "1. User accesses a third-party application -> redirected to the authentication center authorization page. 2. User agrees to authorization -> authentication center returns an authorization code. 3. Third-party application uses the code to obtain an Access Token -> accesses resources with the Token." For the dynamic token (OTP) mechanism in a payment scenario, the corresponding simulation method is: "1. User initiates payment -> server generates a 6-digit OTP (based on time or counter). 2. OTP is sent via SMS / email -> user enters the OTP to complete the payment. 3. Server verifies the validity of the OTP -> executes the deduction operation." The simulated credit indicator information can be various indicator contents after the simulated interaction of the credit token. Simulated credit granting metrics can measure the effectiveness, security, and feasibility of the credit token mechanism in interactive scenarios. For example, simulated credit granting metrics may include: interface TPS, 99th percentile response time, fraudulent request interception rate, false rejection rate, credit granting rate, service availability, and number of circuit breaker triggers.

[0043] In some optional implementations of certain embodiments, after step 104, the steps further include: In response to determining that the above-mentioned interaction scenario is a high-concurrency trust scenario, for each of the at least one trust token mechanism, the following verification steps are performed: Sub-step 1 involves performing stress tests on the credit granting scenario under the aforementioned credit token mechanism to obtain the first verification indicator information. The stress tests include: benchmark testing, end-to-end stress testing, and peak surge testing. In practice, the core of verifying credit granting scenarios under high concurrency is to determine the stability of the credit granting interface under high concurrency (e.g., 100,000+ / second). In practice, high-concurrency credit granting scenarios can be large-scale promotional events. Here, benchmark testing can be using JMeter to simulate single-interface performance. End-to-end stress testing can simulate real traffic distribution by constructing a behavior chain. For example, for large-scale promotional events, the behavior chain can be a user behavior chain. This user behavior chain can be "login -> browse products -> add to cart -> apply for credit -> payment". The real traffic distribution can be the proportion of the payment link. Peak surge testing can simulate sudden traffic spikes (e.g., a sudden increase from 10,000 TPS to 500,000 TPS) to test whether the rate limiting and circuit breaker strategies are effective. The first verification indicator information can be the verification results of the stress tests. The first verification indicator information may include: benchmark test results, end-to-end stress test results, and peak impact test results.

[0044] Sub-step 2 involves dynamically verifying the credit granting capability under the aforementioned credit token mechanism to obtain second verification indicator information. This dynamic credit granting capability verification includes: real-time data linkage testing and inventory-credit synchronization testing. Real-time data linkage testing may involve injecting real-time data (e.g., adding 1000 orders per second) to verify whether the credit limit is dynamically adjusted according to preset rules (e.g., 80% of sales revenue). Inventory-credit synchronization testing may simulate the atomic operation of FBA inventory deduction and credit limit freezing to ensure that the overselling risk is zero. The second verification indicator information can be the verification results of the dynamic credit granting capability verification. This information may include: the test results corresponding to the real-time data linkage test and the test results corresponding to the inventory-credit synchronization test.

[0045] Sub-step 3 involves verifying the risk control rules for the credit granting scenario under the aforementioned credit token mechanism to obtain third verification indicator information. The aforementioned risk control rule verification includes: fraud attack simulation verification and real-time credit score verification. Fraud attack simulation verification can be a verification method that uses chaos engineering tools (such as Chaos Mesh) to inject abnormal requests to simulate fraud attacks. Real-time credit score verification can simulate changes in user behavior data (such as a sudden increase in return rates) and observe whether the model updates the score and adjusts the credit limit within 10 seconds. The third verification indicator information can be the verification results of the risk control rule verification. The third verification indicator information may include: the test results corresponding to the fraud attack simulation verification and the test results corresponding to the real-time credit score verification.

[0046] Sub-step 4 involves performing user experience verification on the credit granting scenario under the aforementioned credit token mechanism to obtain the fourth verification indicator information. This user experience verification includes end-to-end performance verification and abnormal scenario user experience verification. End-to-end performance verification can use RUM (Real User Monitoring) tools to track the complete path time from application to loan disbursement, ensuring that 95% of users wait for ≤2 seconds. Abnormal scenario user experience verification can simulate credit granting failure scenarios (such as insufficient credit limit) to verify the friendliness of the downgrade strategy (such as providing alternative solutions) and whether the churn rate is controlled within 5%. The fourth verification indicator information can be the verification results of the user experience verification. This fourth verification indicator information may include: the test results corresponding to end-to-end performance verification and the test results corresponding to abnormal scenario user experience verification.

[0047] Sub-step 5 involves performing data consistency verification on the credit granting scenario under the aforementioned credit token mechanism to obtain the fifth verification indicator information. This data consistency verification includes: distributed transaction test verification and cache and database synchronization test verification. Distributed transaction test verification can use the Saga pattern to verify the eventual consistency between credit limit deduction and order payment, ensuring automatic compensation in case of service failure. Database synchronization test verification can inject a Redis cluster failure to verify whether the credit data switches to the MySQL backup within 50ms without data loss. The fifth verification indicator information can be the verification results of the data consistency verification. The fifth verification indicator information may include: the test results corresponding to the distributed transaction test verification and the test results corresponding to the database synchronization test verification.

[0048] Sub-step 6: Add the first verification indicator information, the second verification indicator information, the third verification indicator information, the fourth verification indicator information, and the fifth verification indicator information to the corresponding simulated credit granting indicator information.

[0049] Step 105: Based on at least one simulated credit granting indicator information, select a target number of credit granting token mechanisms from the at least one credit granting token mechanism mentioned above, and use them as the primary credit granting token mechanism and the alternative credit granting token mechanism set, respectively.

[0050] In some embodiments, the executing entity can select a target number of credit token mechanisms from the at least one credit token mechanism based on at least one simulated credit indicator information, respectively serving as a primary credit token mechanism and a set of alternative credit token mechanisms. The primary credit token mechanism can be the generated one with the highest adaptability, primarily considering the credit token mechanism to be used. The alternative credit token mechanisms can be those that are no longer used when the primary credit token mechanism has problems. The target number can be a pre-set value. For example, the target number could be 3.

[0051] As an example, the aforementioned execution entity can visualize at least one simulated credit granting indicator, allowing the processing object to select a set of simulated credit granting indicators with high importance. Then, from the at least one credit granting token mechanism, the top-ranked credit granting token mechanisms corresponding to the simulated credit granting indicator set are selected as the primary credit granting token mechanism and the alternative credit granting token mechanism set, respectively. The primary credit granting token mechanism can be the highest-ranked credit granting token mechanism.

[0052] Step 106: Based on the aforementioned primary trust token mechanism and the aforementioned alternative trust token mechanism set, perform the communication interaction processing under the aforementioned interaction scenario.

[0053] In some embodiments, the aforementioned execution entity may perform communication interaction processing in the aforementioned interaction scenario based on the aforementioned primary trust token mechanism and the aforementioned alternative trust token mechanism set.

[0054] As an example, the aforementioned implementing entity can first use the primary trust token mechanism for communication interaction in the interactive scenario. If the primary trust token mechanism is ineffective (i.e., fails to meet the target interaction requirements), a backup trust token mechanism ranked second only to the primary trust token mechanism is selected from the set of backup trust token mechanisms for communication interaction. This process continues, using the next backup trust token mechanism if the current one proves ineffective.

[0055] Step 107: In response to obtaining the actual communication interaction records within the target time period, determine the matching trust token mechanism association information under the above-mentioned interaction scenario, and store the matching trust token mechanism association information and the above-mentioned interaction scenario in the form of key-value pairs in the above-mentioned trust token mechanism association repository.

[0056] In some embodiments, in response to obtaining actual communication interaction records within a target time period, the executing entity may determine the matching trust token mechanism association information for the aforementioned interaction scenario, and store the matching trust token mechanism association information and the aforementioned interaction scenario in key-value pairs to the aforementioned trust token mechanism association repository. The target time period may be a predetermined historical time period prior to the current time. For example, the target time period may be one month prior to the current time. The actual communication interaction records may be records of actual communication interactions with the target external organization. The matching trust token mechanism association information may be the trust token mechanism association information most suitable for use in the interaction scenario.

[0057] In some optional implementations of certain embodiments, after step 107, the steps further include: The first step, in response to the determination that a corresponding trust token mechanism association information exists, is to use the trust token mechanism in the corresponding trust token mechanism association information as the actual trust token mechanism in the aforementioned interaction scenario. The actual trust token mechanism can be the trust token mechanism used to interact with the target external institution in the interaction scenario.

[0058] The second step involves executing the communication interaction processing under the aforementioned interaction scenario based on the actual trust token mechanism described above. The specific implementation will not be elaborated further, but it can be based on the mechanism content corresponding to the actual trust token mechanism for communication interaction processing.

[0059] The third step involves storing the obtained interaction processing results in the actual communication interaction record corresponding to the aforementioned actual trust token mechanism. This allows for mechanism updates to the aforementioned real-time trust token mechanism based on the actual communication interaction record. The interaction processing result can be the communication result after interaction processing within an interaction scenario based on the actual trust token mechanism. Here, updating the actual trust token mechanism using the actual communication interaction record enables the generation of a more accurate and convenient trust token mechanism for communication interaction with the target's external structure in the current interaction scenario, achieving customized processing of the trust token mechanism for different objects and scenarios. In practice, the real-time trust token mechanism can be updated based on a multimodal large model using the actual communication interaction record. This actual communication interaction record can be a record of interactions with the target's external organization.

[0060] It should be noted that the updated trust token mechanism for the target external structure also needs to be stored in the trust token mechanism associated repository for the maintenance of trust token mechanism related information.

[0061] In some optional implementations of certain embodiments, in response to the actual communication interaction records obtained within the target time period, the executing entity can determine the associated information of the matching trust token mechanism under the interaction scenario, including the following steps: The first step, in response to the actual communication interaction records obtained within the target time period, is to execute the second generation step: Sub-step 1: For each of the at least one trust token mechanism mentioned above, select sub-interaction records related to the trust token mechanism from the actual communication interaction records. In practice, the actual communication interaction records can be interaction records of communication interactions using at least one trust token mechanism. The sub-interaction records can be the interaction content of interactions performed under the trust token mechanism.

[0062] Sub-step 2 involves using the aforementioned multimodal large model to extract interaction problem information and interaction advantage information from the aforementioned sub-interaction records. The interaction problem information can be any problems that arise during the target time period when using the trust token mechanism. The interaction advantage information can be any advantages that exist during the target time period when using the trust token mechanism.

[0063] As an example, the aforementioned execution entity can generate information extraction prompts regarding interaction problems and advantages encountered during the use of the trust token mechanism in the sub-interaction records. These prompts are then input into a pre-trained multimodal large model to obtain interaction problem and advantage information.

[0064] Sub-step 3: Based on the above-mentioned interaction problem information and the above-mentioned interaction advantage information, generate the comprehensive interaction score corresponding to the above-mentioned trust token mechanism.

[0065] As an example, firstly, the aforementioned implementing entity can determine the penalty score corresponding to each interaction problem in the interaction problem information. This penalty score can be the impact score of the interaction problem's severity in causing communication interaction issues. That is, the higher the penalty score, the more severe the interaction problem. Next, the advantage reward score corresponding to each interaction advantage in the interaction advantage information is determined. That is, the interaction advantage is the advantage score for using the trust token mechanism for communication interaction. Finally, the obtained penalty scores and advantage reward scores are added together to obtain the overall interaction score.

[0066] The second step involves selecting a matching credit token mechanism from the at least one credit token mechanism described above, where the interaction problem information meets the first condition, the interaction advantage information meets the second condition, and the overall interaction score meets the third condition. The first condition can be that the sum of the problem penalty scores is less than a first preset value. The second condition can be that the sum of the advantage reward scores is greater than a second preset value. The third condition can be that the overall interaction score is greater than a third preset value.

[0067] The third step is to generate the associated information of the matched trust token mechanism based on the trust token generation basis, trust token storage method and security priority information corresponding to the aforementioned matching trust token mechanism.

[0068] As an example, the aforementioned executing entity can combine the credit token generation basis, credit token storage method, and security priority information corresponding to the aforementioned matching credit token mechanism to obtain the credit token mechanism association information.

[0069] In some optional implementations of certain embodiments, after step 107, the steps further include: The first step, in response to the determination that no corresponding credit token mechanism association information exists but a credit mechanism requirement exists, displays a page for filling in the credit token mechanism requirement for the aforementioned target external institution. The credit mechanism requirement can be a requirement for the existence of a credit mechanism usage requirement for the current credit scenario targeting the target external institution. The credit token mechanism requirement filling page can be a page that supports filling in the corresponding mechanism requirement for the credit token mechanism.

[0070] The second step involves responding to the completion of the credit token mechanism time limit requirement information and credit token mechanism security requirement information on the aforementioned credit token mechanism requirement page. This process determines the mechanism feature information set corresponding to the credit token associated mechanism set in the credit token mechanism associated repository. The mechanism feature information includes: time limit features, security features, and at least one scenario feature. The credit token mechanism time limit requirement information can be information specifying the usage time limit and mechanism feedback time limit for the credit token mechanism. For example, the usage time limit is 2 hours, and the corresponding mechanism feedback time limit is that the mechanism feedback time limit must not exceed the target number of seconds. The credit token mechanism security requirement information can be that the security level corresponding to the credit token mechanism is higher than the target security level (i.e., a security level suitable for use in the financial field). Each credit token associated mechanism in the credit token associated mechanism set can be a credit token mechanism stored in the credit token mechanism associated repository. Each credit token associated mechanism has unique corresponding mechanism feature information. The mechanism feature information can be the feature content under each feature corresponding to the credit token mechanism. In practice, multiple features correspond to a credit token mechanism. The time limit feature can be a time limit-level feature corresponding to the credit token mechanism. Security features can be features related to the security level corresponding to the trust token mechanism. Scenario features can be features specific to the interaction scenario. Specifically, at least one scenario feature can be a feature unique to each scenario within the interaction scenario.

[0071] The third step involves setting up a multi-level feature weight set corresponding to the time limit feature, security feature, and at least one scenario feature. The feature weights in this multi-level set are arranged in an arithmetic progression sequence. Each feature weight in the set is greater than 0.6, and the weights form an arithmetic progression.

[0072] Fourth, based on the aforementioned time limit requirements and security requirements of the trust token mechanism, initial feature weights are selected from the aforementioned set of feature weights. These initial feature weights can be initially used to generate the temporary trust token mechanism.

[0073] As an example, the aforementioned executing entity can query the feature weights corresponding to the time limit requirements and security requirements from the association table, using them as initial feature weights. Here, the association table can represent the mapping relationship between the feature content and feature weights corresponding to each feature.

[0074] Fifth, for the initial feature weights, perform the following generation steps: Sub-step 1 involves using a target clustering algorithm to cluster the set of trust token association mechanisms based on the initial feature weights and mechanism feature information set, resulting in a cluster set of trust token association mechanisms. The target clustering algorithm can be a pre-selected algorithm that clusters based on the mechanism features corresponding to the mechanism feature information set. For example, the target clustering algorithm could be the K-means clustering algorithm. Here, regarding the initial feature weights, higher weights are assigned to the feature content corresponding to security features and time-limited features during the clustering process, so that the feature content corresponding to security features and time-limited features is given more importance during clustering.

[0075] Sub-step 2: Select the trust token association mechanism cluster that has the closest relationship with the trust token mechanism time limit requirement information and the trust token mechanism security requirement information from the aforementioned trust token association mechanism cluster set.

[0076] As an example, the aforementioned executing entity can filter out the trust token association mechanism cluster from the authorized token association mechanism cluster set that has the highest degree of content similarity between the corresponding cluster label and the trust token mechanism time limit requirement information and the aforementioned trust token mechanism security requirement information.

[0077] Sub-step 3 involves using the aforementioned multimodal large model to evaluate the scenario applicability of the cluster centers corresponding to the aforementioned trust token association mechanism clusters, thereby obtaining scenario applicability information. The scenario applicability evaluation can be an assessment of whether the cluster centers corresponding to the trust token association mechanism clusters are applicable in interactive scenarios. In practice, scenario applicability information can be in numerical form; the higher the numerical value of the scenario applicability information, the more suitable the mechanism corresponding to the cluster center is for use in interactive scenarios. In practice, scenario applicability information may also include: time limit scores corresponding to time limit features and security scores corresponding to security features.

[0078] As an example, the aforementioned multimodal large model can be used to evaluate the scenario applicability of the cluster center corresponding to the aforementioned trust token association mechanism cluster based on prompt word technology, thereby obtaining scenario applicability information.

[0079] Sub-step 4: In response to determining that the time limit feature and security feature of the above-mentioned scenario-applicable information characterization meet the corresponding requirements, the cluster center corresponding to the above-mentioned trust token association mechanism cluster is used as a temporary trust token mechanism for the above-mentioned target external institution. The scenario-applicable information may include a time limit score corresponding to the time limit feature and a security score corresponding to the security feature. The corresponding requirements may be that the time limit score is higher than a predetermined first score and the security score is higher than a predetermined second score. The temporary trust token mechanism can be a temporarily used trust token mechanism. The first score is used to measure whether the feature content corresponding to the time limit feature meets the requirements of the interaction scenario. The second score is used to measure whether the feature content corresponding to the security feature meets the requirements of the interaction scenario.

[0080] Step 6: In response to the determination that the time limit feature and security feature of the above-mentioned scenario do not meet the corresponding requirements, based on the differences in time limit feature information and security feature information, corresponding feature weights are selected from the feature weight set after removing the initial feature weights, and used as target feature weights. The difference in time limit feature information can be the score difference between the time limit score and the first score. The difference in time limit feature information can characterize whether the temporary trust token mechanism conforms to the interaction scenario from a time limit security perspective. The difference in security feature information can be the score difference between the security score and the second score. The difference in security feature information can characterize whether the temporary trust token mechanism conforms to the interaction scenario from a security perspective.

[0081] As an example, firstly, based on the time-limit difference mapping relationship, the time-limit weight adjustment information corresponding to the aforementioned time-limit feature information differences is determined. The time-limit difference mapping relationship characterizes the mapping relationship between time-limit feature information differences and adjustment weights. Then, based on the security difference mapping relationship, the security weight adjustment information corresponding to the aforementioned security feature information differences is determined. The security difference mapping relationship characterizes the mapping relationship between security feature information differences and adjustment weights. Then, according to the aforementioned time-limit weight adjustment information and the aforementioned security weight adjustment information, the weight values ​​related to each feature in the initial feature weights are adjusted to obtain the adjusted feature weights. Finally, from the feature weight set after removing the initial feature weights, the feature weights whose weight values ​​are closest to the aforementioned adjusted feature weights are selected as the target feature weights.

[0082] Step 7: Use the target feature weights as the initial feature weights and repeat the above generation steps.

[0083] The aforementioned "Steps 1-7" constitute an inventive point of this disclosure, solving another problematic technique: "How to generate a specially customized trust token mechanism in scenarios where there is no corresponding trust token mechanism association information but a trust mechanism requirement exists." Based on this, this disclosure first sets up a trust token mechanism requirement input page to fill in the trust token mechanism time limit requirement information and trust token mechanism security requirement information. Then, it determines the mechanism feature information set corresponding to the trust token mechanism association set in the trust token mechanism association repository, facilitating subsequent clustering processing under corresponding feature weights to query the trust token mechanism association information adapted under the feature requirements. Secondly, it sets a multi-level feature weight set corresponding to the time limit feature, security feature, and at least one scenario feature, wherein the feature weights in the multi-level feature weight set are set according to an arithmetic progression. Here, setting a multi-level feature weight set facilitates the selection of feature weights at each level during the subsequent determination of the temporary trust token mechanism. Next, based on the aforementioned time limit requirements and security requirements of the trust token mechanism, initial feature weights are selected from the aforementioned feature weight set for subsequent clustering processing and the generation of temporary trust token mechanisms. Furthermore, for the initial feature weights, the following generation steps are performed: First, based on the aforementioned initial feature weights and mechanism feature information set, a target clustering algorithm is used to cluster the trust token association mechanism set, resulting in a trust token association mechanism cluster set. Here, clustering allows mechanisms with similar features to be grouped together, facilitating subsequent querying of the most representative trust token association mechanism. Second, from the aforementioned trust token association mechanism cluster set, the trust token association mechanism cluster with the closest correlation to the aforementioned time limit requirements and security requirements of the trust token mechanism is selected to query the most representative trust token association mechanism. Third, using the aforementioned multimodal large model, the scenario applicability assessment is performed on the cluster centers corresponding to the aforementioned trust token association mechanism clusters to obtain scenario applicability information, thereby determining whether the cluster centers support application in interactive scenario content. Fourth, in response to determining that the time limit and security features of the information representation applicable to the above-mentioned scenario meet the corresponding requirements, the cluster center corresponding to the above-mentioned trust token association mechanism cluster is used as a temporary trust token mechanism for the above-mentioned target external institution. Next, in response to determining that the time limit and security features of the information representation applicable to the above-mentioned scenario do not meet the corresponding requirements, based on the differences in time limit feature information and security feature information, corresponding feature weights are selected from the feature weight set after removing the initial feature weights, and used as target feature weights, for further clustering and generation of the temporary trust token mechanism under the target feature weights. Finally, the target feature weights are used as initial feature weights, and the above generation steps are executed again.

[0084] Optionally, after determining that the applicable information representation time limit feature and security feature of the above-mentioned scenario meet the corresponding requirements, and using the cluster center corresponding to the above-mentioned trust token association mechanism cluster as a temporary trust token mechanism for the above-mentioned target external institution, the above-mentioned method further includes: The first step involves obtaining the first actual communication interaction record for the temporary trust token mechanism within the target time period and determining that the token mechanism has a problem. Based on the first actual communication interaction record, a multimodal large model is used to adjust the temporary trust token mechanism, resulting in an adjusted token mechanism. The first actual communication interaction record can be the content of the interaction record during actual communication processing. The target time period can be a period of time after the current time. Within the target time period, the temporary trust token mechanism will be used for communication interaction, resulting in the first actual communication interaction record.

[0085] As an example, firstly, adjustment prompts are generated to modify the temporary trust token mechanism based on the first actual communication interaction record and the question. Then, the adjustment prompts are input into the multimodal large model to obtain the adjusted token mechanism.

[0086] The second step involves generating simulated credit indicator information based on the simulation method corresponding to the aforementioned adjusted token mechanism, simulating the use of the credit token mechanism in the above-mentioned interaction scenario, as the target simulated credit indicator information. Further details are omitted.

[0087] The third step involves, in response to the determination that the aforementioned target simulated credit indicator information meets the corresponding indicator requirements, replacing the temporary credit token mechanism used online with the aforementioned adjustment token mechanism. Further details are omitted.

[0088] Fourth, in response to obtaining the second actual communication interaction record for the adjustment token mechanism within the target time period and confirming that there are no problems with the token mechanism, the trust token mechanism association information for the aforementioned adjustment token mechanism and the aforementioned interaction scenario are stored in the aforementioned trust token mechanism association repository in the form of key-value pairs. Further details are omitted.

[0089] The aforementioned "steps one through four" constitute another inventive aspect of this disclosure. By determining the actual communication interaction records of the temporary trust token mechanism and utilizing a multimodal large model to support mechanism adjustments, the temporary trust token mechanism can be adjusted in real time based on the actual interaction records. Through corresponding simulation methods, it can be determined whether the adjusted token mechanism meets the corresponding indicator requirements, thereby obtaining a more suitable and accurate token mechanism for the interaction scenario.

[0090] The above embodiments of this disclosure have the following beneficial effects: Through the communication interaction method based on a trust token mechanism in some embodiments of this disclosure, a multimodal large model can be used to accurately and efficiently filter out the initial trust token mechanism suitable for the interaction scenario. Based on this, through simulated operation and analysis of real interaction records, the corresponding trust token mechanism for the interaction scenario is further determined. Specifically, the reasons for problems or low efficiency in related communication interactions are: due to the complexity of communication interaction scenarios, different communication interaction scenarios have different communication interaction characteristics. Manually selecting a trust token mechanism is not only inefficient, but the selection of an unsuitable trust token mechanism can lead to communication security or efficiency problems during the communication interaction process. Furthermore, effective control of the trust token mechanism is one of the main research directions at present. Based on this, the communication interaction method based on a trust token mechanism in some embodiments of this disclosure first, in response to receiving interaction information for communication interaction with a target external organization, determines the interaction scenario corresponding to the aforementioned interaction information. Here, by determining the interaction scenario, suitable token mechanism association information for communication interaction with the target external organization can be selected according to the interaction characteristics of the interaction scenario. Then, the system queries the trust token mechanism association repository to determine if there is any trust token mechanism association information corresponding to the aforementioned interaction scenario. This association information includes: the trust token mechanism, the basis for trust token generation, the trust token storage method, and security priority information. By setting up the trust token mechanism association repository, the system can store the association information for various applicable trust token mechanisms under different interaction scenarios, effectively helping to determine the trust token mechanism for each scenario. Furthermore, by setting the trust token mechanism association information to include the trust token mechanism, the basis for trust token generation, the trust token storage method, and security priority information, the system can select the trust token mechanism from multiple perspectives, choosing the most suitable trust token mechanism for the interaction scenario based on matching different characteristics. Next, in response to the determination that no corresponding trust token mechanism association information exists, at least one trust token mechanism for the aforementioned interaction scenario can be accurately generated based on the interaction event information. Here, even when no trust token mechanism association information is found, by fully understanding the interaction characteristics and scenario details through the content of each event in the interaction scenario, at least one trust token mechanism can be accurately preliminarily selected. Next, for each credit token mechanism, based on the simulation method corresponding to the aforementioned credit token mechanism, simulated credit indicator information simulating the use of the credit token mechanism in the aforementioned interaction scenario can be accurately generated, so as to determine the effectiveness and matching of each credit token mechanism through simulated credit indicator information.Furthermore, based on at least one simulated credit granting indicator, a target number of credit token mechanisms with the best matching features are selected from the at least one credit token mechanism mentioned above. These are designated as the primary credit token mechanism and the set of alternative credit token mechanisms, respectively, for subsequent use. Next, based on the primary and alternative credit token mechanisms, communication interaction processing under the aforementioned interaction scenario is executed. Finally, in response to obtaining the actual communication interaction records within the target time period, the associated information of the matched credit token mechanisms under the aforementioned interaction scenario is determined, and the associated information of the matched credit token mechanisms and the aforementioned interaction scenario are stored in the aforementioned credit token mechanism association repository in the form of key-value pairs. This enables the management of the credit token mechanism association information and the selection of credit token mechanisms for subsequent interaction scenarios. In summary, by utilizing a multimodal large model, the appropriate initial trust token mechanism for the interaction scenario can be accurately and efficiently selected. Based on this, through simulation and analysis of real interaction records, the corresponding trust token mechanism for the interaction scenario can be further determined, which can effectively avoid the security and efficiency problems in communication interaction caused by manual selection.

[0091] Further reference Figure 2 As an implementation of the methods shown in the above figures, this disclosure provides some embodiments of a communication interaction device based on a trust token mechanism. These device embodiments are similar to... Figure 1 Corresponding to the method embodiments shown, this communication interaction device based on the trust token mechanism can be specifically applied to various electronic devices.

[0092] like Figure 2As shown, a communication interaction device 200 based on a trust token mechanism includes: a first determining unit 201, a querying unit 202, a first generating unit 203, a second generating unit 204, a filtering unit 205, an execution unit 206, and a second determining unit 206. The first determining unit 201 is configured to determine the interaction scenario corresponding to the received interaction information for communication with a target external organization. The querying unit 202 is configured to query a stored trust token mechanism association repository to determine whether there is trust token mechanism association information corresponding to the interaction scenario. The trust token mechanism association information includes: trust token mechanism, trust token generation basis, trust token storage method, and security priority information. The first generating unit 203 is configured to generate at least one trust token mechanism for the interaction scenario based on various interaction event information in the interaction scenario, in response to determining that no corresponding trust token mechanism association information exists. The second generating unit 204 is configured to, for each of the at least one trust token mechanism, generate a trust token based on the aforementioned trust order. The system generates simulated credit indicator information for the use of the credit token mechanism in the above-mentioned interactive scenario, based on the simulation method corresponding to the credit token mechanism. The filtering unit 205 is configured to filter a target number of credit token mechanisms from the at least one credit token mechanism based on at least one simulated credit indicator information, respectively serving as a primary credit token mechanism and a set of alternative credit token mechanisms. The execution unit 206 is configured to execute communication interaction processing in the above-mentioned interactive scenario based on the primary credit token mechanism and the set of alternative credit token mechanisms. The second determining unit 207 is configured to, in response to the actual communication interaction records obtained within the target time period, determine the credit token mechanism association information matched in the above-mentioned interactive scenario, and store the matched credit token mechanism association information and the above-mentioned interactive scenario in key-value pairs in the credit token mechanism association repository.

[0093] It is understandable that the units described in the communication interaction device 200 based on the trust token mechanism are related to the reference. Figure 1 The steps in the described method correspond to each other. Therefore, the operations, features, and beneficial effects described above for the method also apply to the communication interaction device 200 based on the trust token mechanism and the units contained therein, and will not be repeated here.

[0094] The following is for reference. Figure 3 It shows a schematic diagram of the structure of an electronic device (e.g., an electronic device) 300 suitable for implementing some embodiments of the present disclosure. Figure 3 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments of this disclosure.

[0095] like Figure 3 As shown, the electronic device 300 may include a processing unit (e.g., a central processing unit, a graphics processing unit, etc.) 301, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 302 or a program loaded from a storage device 308 into a random access memory (RAM) 303. The RAM 303 also stores various programs and data required for the operation of the electronic device 300. The processing unit 301, ROM 302, and RAM 303 are interconnected via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0096] Typically, the following devices can be connected to I / O interface 305: input devices 306 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 307 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 308 including, for example, magnetic tapes, hard disks, etc.; and communication devices 309. Communication device 309 allows electronic device 300 to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 3 An electronic device 300 with various devices is shown; however, it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed alternatively. Figure 3 Each box shown can represent a device or multiple devices as needed.

[0097] In particular, according to some embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, some embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication device 309, or installed from storage device 308, or installed from ROM 302. When the computer program is executed by processing device 301, it performs the functions defined in the methods of some embodiments of this disclosure.

[0098] It should be noted that, in some embodiments of this disclosure, the computer-readable medium described above may be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In some embodiments of this disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In some embodiments of this disclosure, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wires, optical fibers, RF (radio frequency), etc., or any suitable combination thereof.

[0099] In some implementations, clients and servers can communicate using any currently known or future-developed network protocol such as HTTP (Hypertext Transfer Protocol) and can interconnect with digital data communication (e.g., communication networks) of any form or medium. Examples of communication networks include local area networks (“LANs”), wide area networks (“WANs”), the Internet (e.g., the Internet of Things), and peer-to-peer networks (e.g., ad hoc peer-to-peer networks), as well as any currently known or future-developed networks.

[0100] The aforementioned computer-readable medium may be included in the aforementioned electronic device; or it may exist independently and not assembled into the electronic device. The aforementioned computer-readable medium carries one or more programs, which, when executed by the electronic device, cause the electronic device to: in response to receiving interaction information for communication with a target external organization, determine the interaction scenario corresponding to the interaction information; query a stored trust token mechanism association repository to determine whether trust token mechanism association information corresponding to the interaction scenario exists, wherein the trust token mechanism association information includes: trust token mechanism, trust token generation basis, trust token storage method, and security priority information; in response to determining that no corresponding trust token mechanism association information exists, generate at least one trust token mechanism for the interaction scenario based on various interaction event information under the interaction scenario; and for each of the at least one trust token mechanism… The trust token mechanism generates simulated trust indicator information based on the simulation method corresponding to the aforementioned trust token mechanism, simulating the use of the trust token mechanism in the aforementioned interaction scenario. Based on at least one simulated trust indicator information, a target number of trust token mechanisms are selected from the at least one trust token mechanism, respectively serving as a primary trust token mechanism and a set of alternative trust token mechanisms. Based on the primary trust token mechanism and the set of alternative trust token mechanisms, communication interaction processing in the aforementioned interaction scenario is executed. In response to obtaining the actual communication interaction records within a target time period, the associated information of the trust token mechanism matched in the aforementioned interaction scenario is determined, and the associated information of the matched trust token mechanism and the aforementioned interaction scenario are stored in the aforementioned trust token mechanism associated repository in the form of key-value pairs.

[0101] Computer program code for performing operations of some embodiments of this disclosure can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0102] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0103] The units described in some embodiments of this disclosure can be implemented in software or hardware. The described units can also be housed in a processor; for example, a processor can be described as including a first determining unit, a querying unit, a first generating unit, a second generating unit, a filtering unit, an execution unit, and a second determining unit. The names of these units do not necessarily limit the specific unit; for example, the first determining unit can also be described as "a unit that, in response to receiving interactive information for communication with a target external organization, determines the interactive scenario corresponding to the interactive information."

[0104] The functions described above in this document can be performed at least in part by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), system-on-a-chip (SoCs), complex programmable logic devices (CPLDs), and so on.

[0105] The above description is merely a selection of preferred embodiments of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in the embodiments of this disclosure is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described inventive concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features with similar functions disclosed in the embodiments of this disclosure.

Claims

1. A communication interaction method based on a trust token mechanism, comprising: In response to receiving interaction information for communication with a target external organization, the interaction scenario corresponding to the interaction information is determined; The system queries the trust token mechanism associated repository to determine if there is trust token mechanism associated information corresponding to the interaction scenario. The trust token mechanism associated information includes: trust token mechanism, trust token generation basis, trust token storage method, security priority information, and interaction scenario. In response to determining that there is no corresponding trust token mechanism associated information, at least one trust token mechanism for the interaction scenario is generated based on the interaction event information in the interaction scenario and using a multimodal large model. For each credit token mechanism, based on the simulation method corresponding to the credit token mechanism, simulated credit indicator information is generated to simulate the use of the credit token mechanism in the interaction scenario. Based on at least one simulated credit granting indicator information, a target number of credit granting token mechanisms are selected from the at least one credit granting token mechanism, which are respectively used as the primary credit granting token mechanism and the alternative credit granting token mechanism set; Based on the primary trust token mechanism and the alternative trust token mechanism set, perform the communication interaction processing in the interaction scenario; In response to obtaining the actual communication interaction records within the target time period, the system determines the matching trust token mechanism association information for the interaction scenario, and stores the matching trust token mechanism association information and the interaction scenario in the form of key-value pairs in the trust token mechanism association repository.

2. The method according to claim 1, wherein, The step of generating at least one authorization token mechanism for the interaction scenario based on the interaction event information in the interaction scenario and using a multimodal large model includes: Based on the information of each interactive event, an interactive knowledge graph with multiple layers of content superimposed and nested is generated. The interactive knowledge graph includes: multiple interactive event chains, each interactive event chain has multiple connected branches, and each branch is supplementary content in the process of interacting with the interactive event chain. Retrieve information sets of credit granting events and classic case information related to the interaction scenario from an external knowledge base; Based on the interaction knowledge graph, the information of each interaction event, the set of trust event information, and the set of classic case information for the scenario, at least one trust token mechanism is generated under the interaction scenario using the multimodal large model.

3. The method according to claim 2, wherein, The step of generating at least one trust token mechanism for the interaction scenario based on the interaction knowledge graph, the information of each interaction event, the set of trust event information, and the set of classic case information, using the multimodal large model, includes: Generate prompt information for the authorization token mechanism that generates authorization tokens for the interaction knowledge graph, the information of each interaction event, the authorization event information set, and the information set of classic case scenarios; The authorization token mechanism generates a prompt message which is then input into the multimodal large model to obtain at least one first initial authorization token mechanism. For at least one first initial trust token mechanism, perform the following first generation step: Generate verification prompt information for the feasibility, security and matching verification of the at least one first initial trust token mechanism; The verification prompt information is input into a large model used for querying vulnerabilities in the trust token mechanism to obtain at least one vulnerability information in the trust token mechanism. In response to determining that the vulnerability type corresponding to at least one trust token mechanism vulnerability information is not in a predetermined vulnerability type set and the vulnerability level is lower than the target level, at least one first initial trust token mechanism is determined as at least one trust token mechanism.

4. The method according to claim 3, wherein, The method further includes: In response to determining that at least one trust token mechanism vulnerability information contains a target trust token mechanism vulnerability information whose corresponding vulnerability type is in a predetermined vulnerability type set and / or whose vulnerability level is not less than the target level, at least one target trust token mechanism vulnerability information and the corresponding at least one first initial trust token mechanism are added to the trust token mechanism to generate a prompt message, and an addition prompt message is obtained; The added prompt information is input into the multimodal large model to obtain at least one second initial trust token mechanism; The first generation step is performed by using the at least one second initial trust token mechanism as at least one first initial trust token mechanism.

5. The method according to claim 1, wherein, The method further includes: In response to determining that there is a corresponding credit token mechanism association information, the credit token mechanism in the corresponding credit token mechanism association information is used as the actual credit token mechanism in the interaction scenario; Based on the actual authorization token mechanism, perform the communication interaction processing in the interaction scenario; The obtained interaction processing results are stored in the actual communication interaction record corresponding to the actual trust token mechanism, so as to update the real-time trust token mechanism according to the actual communication interaction record corresponding to the trust token mechanism.

6. The method according to claim 1, wherein, After generating simulated credit indicator information for each credit token mechanism based on the simulation method corresponding to the credit token mechanism, simulating the use of the credit token mechanism in the interaction scenario, the method further includes: In response to determining that the interaction scenario is a high-concurrency trust scenario, for each of the at least one trust token mechanism, the following verification steps are performed: The credit granting scenario under the credit granting token mechanism is subjected to stress testing to verify the first verification indicator information. The stress testing verification includes: benchmark testing, end-to-end stress testing, and peak impact testing. Dynamic credit granting capability verification is performed on the credit granting scenario under the credit granting token mechanism to obtain second verification indicator information. The dynamic credit granting capability verification includes: real-time data linkage test and inventory and credit granting synchronization test. The credit granting scenario under the credit granting token mechanism is subjected to risk control rule verification to obtain third verification indicator information, wherein the risk control rule verification includes: fraud attack simulation verification and credit score real-time verification. User experience verification is performed on the credit granting scenario under the credit granting token mechanism to obtain fourth verification indicator information, wherein the user experience verification includes: end-to-end performance verification and abnormal scenario user experience verification; Data consistency verification is performed on the credit granting scenario under the credit granting token mechanism to obtain the fifth verification indicator information, wherein the data consistency verification includes: distributed transaction test verification and cache and database synchronization test verification; Add the first verification indicator information, the second verification indicator information, the third verification indicator information, the fourth verification indicator information, and the fifth verification indicator information to the corresponding simulated credit granting indicator information.

7. The method according to claim 1, wherein, The response to the actual communication interaction records within the obtained target time period determines the associated information of the matching trust token mechanism in the interaction scenario, including: In response to the actual communication interaction records obtained within the target time period, the second generation step is executed: For each of the at least one trust token mechanism, select sub-interaction records related to the trust token mechanism from the actual communication interaction records; Using the multimodal large model, the interaction problem information and interaction advantage information in the sub-interaction records are extracted; Based on the interaction problem information and the interaction advantage information, a comprehensive interaction score corresponding to the trust token mechanism is generated; From the at least one credit token mechanism, select the credit token mechanism in which the interaction question information satisfies the first condition, the interaction advantage information satisfies the second condition, and the interaction comprehensive score satisfies the third condition, and use it as the matching credit token mechanism. Based on the credit token generation basis, credit token storage method, and security priority information corresponding to the matching credit token mechanism, generate the associated information of the matching credit token mechanism.

8. A communication interaction device based on a trust token mechanism, comprising: The first determining unit is configured to determine the interaction scenario corresponding to the interaction information in response to receiving interaction information that communicates with an external target mechanism. The query unit is configured to query the trust token mechanism association repository to see if there is trust token mechanism association information corresponding to the interaction scenario. The trust token mechanism association information includes: trust token mechanism, trust token generation basis, trust token storage method and security focus information. The first generation unit is configured to, in response to determining that there is no corresponding trust token mechanism associated information, generate at least one trust token mechanism in the interaction scenario based on the interaction event information in the interaction scenario and using a multimodal large model. The second generation unit is configured to generate simulated credit indicator information for each credit token mechanism, based on the simulation method corresponding to the credit token mechanism, to simulate the use of the credit token mechanism in the interaction scenario. The filtering unit is configured to select a target number of credit token mechanisms from the at least one credit token mechanism based on at least one simulated credit indicator information, and to serve as a primary credit token mechanism and a set of alternative credit token mechanisms, respectively. The execution unit is configured to perform communication interaction processing in the interaction scenario according to the primary trust token mechanism and the alternative trust token mechanism set; The second determining unit is configured to, in response to the actual communication interaction records obtained within the target time period, determine the matching trust token mechanism association information under the interaction scenario, and store the matching trust token mechanism association information and the interaction scenario in the form of key-value pairs in the trust token mechanism association repository.

9. An electronic device, comprising: One or more processors; Storage device, on which one or more programs are stored, When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in any one of claims 1-7.

10. A computer-readable medium having a computer program stored thereon, wherein, When the program is executed by the processor, it implements the method as described in any one of claims 1-7.

Images