Cross-domain model communication method, apparatus and device, and computer readable storage medium

Abstract

The invention relates to the technical field of data transmission, and discloses a cross-domain model communication method, device and equipment and a computer readable storage medium. The method is applied to a cross-domain model communication system, the system comprises a user domain, a data domain and a gatekeeper, and the method comprises the following steps: sending request information of a model in the data domain to the gatekeeper through the user domain, and performing security auditing on the request information through the gatekeeper; when the security check is passed, sending the request information to a data domain through a gatekeeper, and generating response information based on the request information through a model; sending the response information to a gatekeeper through the data domain, and performing security auditing on the response information through the gatekeeper; and when the security check is passed, the response information is sent to the user domain through the gatekeeper. When the model carries out data transmission between the user domain and the data domain, the transmitted data is subjected to security auditing through the gatekeeper, and the data transmission is carried out only when the security auditing is passed, so that the security of cross-domain communication of the model is improved.

Description

Technical Field

[0001] This application relates to the field of data transmission technology, and in particular to a cross-domain model communication method, apparatus, device, and computer-readable storage medium. Background Technology

[0002] In scenarios involving cross-domain communication of large language models, long-lived cross-domain connections are achieved by directly binding the data domain and the user domain, ensuring the timeliness of large language model communication. However, the data domain and the user domain have different security mechanisms. Long-lived cross-domain communication breaks these security mechanisms, allowing threats from the user domain to infiltrate the data domain, and sensitive information from the data domain to leak out of the user domain without control. This results in the inability to guarantee the security of cross-domain communication of large language models. Summary of the Invention

[0003] In view of this, the purpose of this application is to overcome the shortcomings of the prior art and provide a cross-domain model communication method. The method is applied to a cross-domain model communication system, which includes a user domain, a data domain, and a gateway. The method includes: The user domain sends the request information for the model in the data domain to the gateway, and the gateway performs security audit on the request information. Upon successful security audit, the request information is sent to the data domain via the gateway, and the model generates response information based on the request information. The response information is sent to the gateway through the data domain, and the gateway performs security audit on the response information. Upon successful security audit, the response information is sent to the user domain via the gateway.

[0004] In one embodiment, the user domain includes a client and a front-end proxy. The step of sending request information for a model in the data domain to the gateway through the user domain includes: The client generates an interaction request for the model in the data domain and sends the interaction request to the front-end agent. The interaction request is encapsulated by the front-end proxy to generate request information with target identifier and interaction request content, and the request information is sent to the gateway.

[0005] In one embodiment, the data domain includes a caching layer and a back-end proxy. The step of sending the request information to the data domain through the gateway and generating response information based on the request information using the model includes: The request information is sent to the back-end proxy through the network gateway, and the back-end proxy parses the request information to obtain the interaction request content and the target identifier; The back-end proxy sends the interaction request content and the target identifier to the model, and the model generates response information based on the interaction request content. The model caches the response information into the cache layer based on the target identifier.

[0006] In one embodiment, the step of sending the response information to the gateway through the data domain includes: The response information is split according to a preset splitting rule by the back-end proxy to obtain split response data; According to the splitting order, each splitting response data is sent to the gateway in sequence.

[0007] In one embodiment, the step of sending the response information to the user domain through the gateway includes: The gateway sends each of the split response data to the front-end agent sequentially based on the receiving order; The front-end agent sequentially sends each of the split response data to the client.

[0008] In one embodiment, the method further includes: In response to a query from the user domain, the data domain extracts the target data corresponding to the query from the cache layer of the data domain and sends the target data to the gateway. The target data is subjected to security audit through the network gateway, and when the security audit is passed, the target data is sent to the user domain through the network gateway.

[0009] In one embodiment, the method further includes: The target identifier is determined based on the receiving completion information or cache expiration instruction sent by the user domain in response to the data domain. Clear the cached data corresponding to the target identifier in the cache layer.

[0010] This application also provides a cross-domain model communication device, the cross-domain model communication device comprising: The first sending module is used to send request information for a model in the data domain to the network gateway through the user domain, and to perform security audit on the request information through the network gateway. The first review module is used to send the request information to the data domain through the network gateway when the security review is passed, and to generate response information based on the request information through the model. The second audit module is used to send the response information to the gateway through the data domain, and to perform security audit on the response information through the gateway. The second sending module is used to send the response information to the user domain through the network gateway when the security audit is passed.

[0011] This application also provides a computer device, which includes a processor and a memory, the memory storing a computer program, and the processor executing the computer program to implement the above-described cross-domain model communication method.

[0012] This application also provides a computer-readable storage medium storing a computer program that, when run on a processor, executes the above-described cross-domain model communication method.

[0013] The embodiments of this application have the following beneficial effects: In this embodiment, a request for a model in the data domain is sent from the user domain to a network gateway, which performs a security audit on the request. If the security audit is successful, the request is sent to the data domain via the network gateway, and the model generates a response based on the request. The response is then sent from the data domain to the network gateway, which performs a security audit on the response. If the security audit is successful, the response is sent to the user domain via the network gateway. When the model transmits data between the user domain and the data domain, the network gateway performs a security audit on the transmitted data. Data transmission only occurs after the security audit is successful, preventing threats from the user domain from penetrating the data domain and preventing sensitive information from the data domain from leaking out of the user domain without control, thus improving the security of cross-domain communication between the model and the network gateway. Attached Figure Description

[0014] To more clearly illustrate the technical solutions of this application, the accompanying drawings used in the embodiments will be briefly described below. It should be understood that the following drawings only show some embodiments of this application and therefore should not be considered as a limitation on the scope of protection of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0015] Figure 1 A flowchart illustrating the first embodiment of the cross-domain model communication method provided in this application; Figure 2 A flowchart illustrating a second embodiment of the cross-domain model communication method provided in this application; Figure 3 This is a schematic diagram of the cross-domain model communication system provided in this application; Figure 4 A flowchart illustrating the third embodiment of the cross-domain model communication method provided in this application; Figure 5 A flowchart illustrating the fourth embodiment of the cross-domain model communication method provided in this application; Figure 6 A flowchart illustrating the fifth embodiment of the cross-domain model communication method provided in this application; Figure 7 This is a schematic diagram of the cross-domain model communication device provided in this application. Detailed Implementation

[0016] The technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments.

[0017] The components of the embodiments of this application described and illustrated in the accompanying drawings can be arranged and designed in a variety of different configurations. Therefore, the following detailed description of the embodiments of this application provided in the drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of the application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0018] In the following, the terms “comprising,” “having,” and their cognates, which may be used in various embodiments of this application, are intended only to indicate a particular feature, number, step, operation, element, component, or combination thereof, and should not be construed as excluding, firstly, the presence of one or more other features, numbers, steps, operations, elements, components, or combinations thereof, or adding the possibility of one or more features, numbers, steps, operations, elements, components, or combinations thereof.

[0019] Furthermore, the terms "first," "second," and "third" are used only to distinguish descriptions and should not be interpreted as indicating or implying relative importance.

[0020] Unless otherwise specified, all terms used herein (including technical and scientific terms) shall have the same meaning as commonly understood by one of ordinary skill in the art to which the various embodiments of this application pertain. Terms (such as those defined in commonly used dictionaries) shall be interpreted as having the same meaning as in their contextual meaning in the relevant technical field and shall not be construed as having an idealized or overly formal meaning, unless clearly defined in the various embodiments of this application.

[0021] It is understood that the method in this application is applied to cross-domain model communication systems, which include user domains, data domains, and gateways. The user domain refers to the network environment where the user terminal resides, such as an office network, the internet, or a low-security network. Users initiate service requests to the model from this location, where the security level is relatively low and they may face more external threats. The data domain refers to the network where the core system carrying the model and its sensitive data resides, such as a high-security intranet, data center, or dedicated AI computing cluster. It stores critical assets such as model weights, training data, and context states, and has a higher security level, requiring strict prevention of unauthorized access and data leakage. A gateway is a security device used to achieve physical or logical isolation between networks of different security levels while allowing controlled data exchange.

[0022] The following detailed description of some embodiments of this application is provided in conjunction with the accompanying drawings. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0023] Please refer to Figure 1 , Figure 1 This is a flowchart illustrating a first embodiment of the cross-domain model communication method provided in this application. The method is applied to a cross-domain model communication system, which includes a user domain, a data domain, and a network gateway. The method includes: Step S101: Send the request information for the model in the data domain to the gateway through the user domain, and perform security audit on the request information through the gateway.

[0024] In this embodiment, when a user initiates a request for information about a model in the data domain from a user terminal in the user domain, the request information is sent to the network gateway through the user domain, and the network gateway performs security audit on the request information.

[0025] It should be noted that the user domain and the data domain are two network environments with different security levels. Data transmitted between the user domain and the data domain must be subject to security audits by the network gateway to ensure that threats from the user domain do not infiltrate the data domain and that sensitive information from the data domain does not leak out of the user domain without control.

[0026] In one embodiment, the model in the data domain is a large language model, such as ChatGPT or deepseek. The user terminal includes a client that interacts with the large language model.

[0027] Step S102: When the security audit is passed, the request information is sent to the data domain through the network gateway, and the response information is generated based on the request information through the model.

[0028] In this embodiment, when the gateway performs a security audit on the request information and determines that the security audit has passed, the request information is sent to the data domain through the gateway, and a response information is generated based on the request information by the model. It should be noted that the gateway can perform a security audit on information sent from the user domain to the data domain. The audit content usually involves integrity verification, protocol compliance checks, and security policy matching of data packets from the user domain.

[0029] Step S103: Send the response information to the gateway through the data domain, and perform security audit on the response information through the gateway.

[0030] In this embodiment, after the model generates response information based on the request information, the response information is sent to the network gateway through the data domain, and the network gateway performs security audits on the response information. It should be noted that the network gateway can perform security audits on information sent from the data domain to the user domain. The audit content typically includes: protocol compliance audit, data content security review, access control and auditing, etc.

[0031] Step S104: When the security audit is passed, the response information is sent to the user domain through the gateway.

[0032] In this embodiment, when the gateway performs security audit on the response information and the security audit is passed, the response information is sent to the user domain through the gateway. The user terminal in the user domain can receive the response information sent by the data domain, and the user terminal can display the response information to the user through the client.

[0033] This embodiment of the cross-domain model communication system sends request information for models in the data domain to a network gateway through the user domain, and the network gateway performs security audits on the request information. If the security audit is successful, the request information is sent to the data domain through the network gateway, and the model generates response information based on the request information. The response information is then sent to the network gateway through the data domain, and the network gateway performs security audits on the response information. If the security audit is successful, the response information is sent to the user domain through the network gateway. When models transmit data between the user domain and the data domain, the network gateway performs security audits on the transmitted data. Data transmission only occurs after the security audit is successful, preventing threats from the user domain from infiltrating the data domain and preventing sensitive information from the data domain from leaking out of the user domain without control, thus improving the security of cross-domain model communication.

[0034] Please refer to Figure 2 , Figure 2 This is a flowchart illustrating a second embodiment of the cross-domain model communication method provided in this application. The difference between the second embodiment and the first embodiment is that the user domain includes a client and a front-end proxy. The step of sending request information for a model in the data domain to the gateway through the user domain includes: Step S201: Generate an interaction request for the model in the data domain through the client, and send the interaction request to the front-end agent.

[0035] In this embodiment, as Figure 3 As shown, Figure 3 This is a schematic diagram of the cross-domain model communication system provided in this application. The user domain of the cross-domain model communication system includes a client and a front-end agent, and a long-lived communication connection is established between the client and the front-end agent. When a user initiates an interaction request to a model in the data domain through the client in the user domain, the interaction request is sent to the front-end agent through the client.

[0036] Step S202: The interaction request is encapsulated by the front-end proxy to generate request information with target identifier and interaction request content, and the request information is sent to the gateway.

[0037] In this embodiment, after receiving an interaction request from the client, the front-end proxy encapsulates the interaction request to generate request information containing a target identifier and the interaction request content, and then sends the request information to the network gateway. It can be understood that the front-end proxy encapsulates the interaction request according to preset encapsulation rules, generating request information containing a target identifier and the interaction request content.

[0038] In one embodiment, the model in the data domain is a large language model, and the target identifier is used to describe the dialogue group corresponding to the interaction request issued by the client. The interaction request content refers to the question raised by the client to the large language model. For example, in the first round: the user asks "What is blockchain?", the front-end agent assigns the target identifier requestId=A to this dialogue group during encapsulation, and the interaction request content is "What is blockchain?"; in the second round: the user asks "Can you give an example?", since this is the same dialogue group, the front-end agent reuses the target identifier requestId=A for this dialogue group during encapsulation, and the interaction request content is "Can you give an example?".

[0039] The cross-domain model communication system in this embodiment generates an interaction request to a model in the data domain through a client, and sends the interaction request to a front-end agent. The front-end agent encapsulates the interaction request, generates request information with a target identifier and the interaction request content, and sends the request information to the network gateway. Establishing a long-lived communication connection between the client and the front-end agent in the user domain improves the data transmission efficiency between the client and the front-end agent, thus helping to improve the efficiency of cross-domain model communication.

[0040] Please refer to Figure 4 , Figure 4This is a flowchart illustrating a third embodiment of the cross-domain model communication method provided in this application. The difference between this third embodiment and the first to second embodiments is that the data domain includes a cache layer and a post-proxygen proxy. The step of sending the request information to the data domain through the gateway and generating response information based on the request information through the model includes: Step S301: The request information is sent to the back-end proxy through the network gateway, and the back-end proxy parses the request information to obtain the interaction request content and the target identifier.

[0041] In this embodiment, as Figure 3 As shown, the data domain of the cross-domain model communication system includes a caching layer, a back-end proxy, and a Large Language Model (LLM). A long-lived communication connection is established between the caching layer, the back-end proxy, and the LLM. When the network gateway performs security audits on the request information and determines that the security audit has passed, the request information is sent to the back-end proxy through the network gateway. The back-end proxy then parses the request information to obtain the interactive request content and target identifier.

[0042] Step S302: The interaction request content and the target identifier are sent to the model through the back-end proxy, and the model generates response information based on the interaction request content.

[0043] In this embodiment, the interaction request content and target identifier are sent to the model through a back-end proxy. The model generates response information based on the interaction request content, and the response information generated by the model is bound to the target identifier.

[0044] Step S303: The response information is cached in the cache layer based on the target identifier by the model.

[0045] In this embodiment, the response information is cached in the cache layer based on the target identifier by the model. It is understood that, as... Figure 3 As shown, the caching layer allocates different cache locations to store corresponding response information based on different target identifiers. For example, requestId1, requestId2, requestId3, and requestId4 in the diagram represent different identifiers corresponding to different dialog groups. If a target identifier is determined and compared with other identifiers already stored in the caching layer, and a match is found, the response information corresponding to the target identifier is stored in the cache location corresponding to that identifier. If no match is found, a cache location belonging to the target identifier is allocated in the caching layer, and the response information corresponding to the target identifier is then stored in that cache location.

[0046] Understandably, since user domains and data domains cannot directly transmit data, caching the model-generated response information in a cache layer helps ensure that the response information is not easily lost and improves the accuracy of subsequent data transmission.

[0047] Please refer to Figure 5 , Figure 5 This is a flowchart illustrating the fourth embodiment of the cross-domain model communication method provided in this application. The difference between the fourth embodiment and the first to third embodiments is that the step of sending the response information to the gateway through the data domain includes: Step S401: The response information is split according to a preset splitting rule by the back-end agent to obtain split response data.

[0048] In this embodiment, the response information is split according to a preset splitting rule by a post-proximity proxy to obtain split response data. It can be understood that the preset splitting rule is to split the response information according to the user's reading order and the paragraphs or sentences of the response information. For example, the response information contains two paragraphs, each containing three sentences; optionally, the post-proximity proxy can split the first paragraph of the text information into one split response data, and the second paragraph of the text information into one split response data; optionally, the post-proximity proxy can split each sentence in the first paragraph of the text information into three split response data, and each sentence in the second paragraph of the text information into three split response data.

[0049] Step S402: In accordance with the splitting order, each splitting response data is sent to the network gateway in sequence.

[0050] In this embodiment, the back-end agent sends each split response data to the network gateway in the order of splitting. It should be noted that since the response information is split according to the order in which the user reads the response information, the back-end agent sends each split response data to the network gateway after obtaining it in the order in which the user reads the response information.

[0051] In one embodiment, the step of sending the response information to the user domain through the gateway includes: Step S403: Based on the receiving order, the network gateway sequentially sends each of the split response data to the front-end agent.

[0052] In this embodiment, after receiving the split response data, the network gateway performs a security audit on the split response data. If the security audit is successful, the network gateway sends each split response data to the front-end agent sequentially based on the receiving order. It can be understood that the back-end agent, following the user's reading order of the response information, splits each split response data and sends it to the network gateway. At this point, the network gateway performs a security audit on the split response data and sequentially sends each split response data that passes the security audit to the front-end agent.

[0053] Step S404: The front-end agent sequentially sends each of the split response data to the client.

[0054] In this embodiment, each time the front-end agent receives a split response data, it parses the split response data accordingly, obtains the parsed return result, and then sends the parsed return result to the client. This allows the client to gradually display the response information according to the user's reading order, thereby improving the coherence and timeliness of communication in the large language model.

[0055] It should be noted that in scenarios involving cross-domain communication of large language models, the user domain and data domain are usually located in network environments with different security levels, requiring a security isolation mechanism to achieve data interaction. Existing technologies often rely on file transfer and simple packet forwarding for cross-domain data output, which is difficult to adapt to the real-time bidirectional communication requirements of large language models, resulting in insufficient timeliness and contextual coherence in model communication. This embodiment, however, utilizes the collaborative operation of a back-end proxy in the data domain, a network gateway, a front-end proxy in the user domain, and the client to segment the model's response information, achieving streaming data output with simultaneous generation, segmentation, and transmission. The network gateway performs security audits, while the front-end proxy implements asynchronous data push with simultaneous reception and transmission. Without compromising security isolation, this allows cross-domain model communication to approach the effect of long-connection communication. Each response fragment generated by the model can be promptly transmitted back to the client, avoiding the significant delays caused by the traditional "whole-segment waiting" mode. Users experience a low-latency, continuous output experience similar to a direct-connection LLM, significantly improving the quality of human-computer interaction. This, in turn, improves the coherence and timeliness of cross-domain model communication.

[0056] Please refer to Figure 6 , Figure 6 This is a flowchart illustrating the fifth embodiment of the cross-domain model communication method provided in this application. The difference between the fifth embodiment and the first to fourth embodiments is that the method further includes: Step S501: In response to the query information sent by the user domain, the data domain extracts the target data corresponding to the query information from the cache layer of the data domain and sends the target data to the network gateway.

[0057] Step S502: Perform security audit on the target data through the network gateway, and when the security audit is passed, send the target data to the user domain through the network gateway.

[0058] In this embodiment, when a client in the user domain receives response information from the model in the data domain, it verifies the completeness of the received response information. If it determines that a segment is missing, it sends query information to the data domain. The query information includes the target identifier of the dialog group and the missing content. The data domain responds to the query information sent by the user domain by retrieving the target data corresponding to the query information from its cache layer and sends the target data to the network gateway. The network gateway performs security audits on the target data, and if the security audit is successful, the target data is sent to the user domain through the network gateway.

[0059] It should be noted that the cache layer in the data domain caches the response information corresponding to the target identifiers of different dialogue groups output by the model. The cache layer can find the corresponding response information based on the target identifiers of the dialogue groups in the query information, and then extract the target data from the response information based on the missing content in the query information.

[0060] In this embodiment, when the client in the user domain detects that there are missing segments in the received response information, it can send query information to the user domain to extract the missing target data in the cache layer of the user domain and return it to the client, which helps to improve the accuracy of data transmission when the model communicates across network domains.

[0061] In one embodiment, the method further includes: Step S503: In response to the reception completion information sent by the user domain or in response to the cache expiration instruction, the data domain determines the target identifier based on the reception completion information or the cache expiration instruction.

[0062] Step S504: Clear the cached data corresponding to the target identifier in the cache layer.

[0063] In this embodiment, the data domain responds to a reception completion message sent by the user domain or to a cache expiration instruction. A target identifier is determined based on the reception completion message or the cache expiration instruction. The cached data corresponding to the target identifier in the cache layer is then cleared. In one embodiment, the user domain verifies the integrity of the received response information. If it determines that no fragments are missing, it sends a reception completion message to the data domain, which carries the target identifier. The data domain determines the target identifier based on the reception completion message and clears the cached data corresponding to the target identifier in the cache layer. In another embodiment, each target identifier in the cache layer has a set cache expiration period. A timer in the cache layer continuously estimates the remaining cache expiration period for each target identifier. When the remaining cache expiration period for a particular target identifier reaches zero, a cache expiration instruction is triggered. This cache expiration instruction carries the target identifier. The data domain determines the target identifier based on the reception completion message and clears the cached data corresponding to the target identifier in the cache layer.

[0064] This embodiment can clear the corresponding cached data in the cache layer when it is determined that the user domain has fully received the response information from the model, or when the cached data reaches its cache expiration date. This avoids redundancy in the cached data and helps improve the utilization rate of the cache layer.

[0065] refer to Figure 7 , Figure 7 This is a schematic diagram of the cross-domain model communication device provided in this application. The cross-domain model communication device includes: The first sending module 10 is used to send request information for a model in the data domain to the network gateway through the user domain, and to perform security audit on the request information through the network gateway. The first audit module 20 is used to send the request information to the data domain through the network gateway when the security audit is passed, and to generate response information based on the request information through the model. The second audit module 30 is used to send the response information to the network gateway through the data domain, and to perform security audit on the response information through the network gateway; The second sending module 40 is used to send the response information to the user domain through the network gateway when the security audit is passed.

[0066] It is understood that the cross-domain model communication device in this embodiment corresponds to the cross-domain model communication method in the above embodiment. The options in the above embodiment are also applicable to this embodiment, so they will not be described again here.

[0067] This application also provides a computer device, exemplary of which includes a processor and a memory, wherein the memory stores a computer program, and the processor executes the computer device to perform the above-described cross-domain model communication method by running the computer program.

[0068] The processor can be an integrated circuit chip with signal processing capabilities. The processor can be a general-purpose processor, including at least one of a Central Processing Unit (CPU), Graphics Processing Unit (GPU), Network Processor (NP), Digital Signal Processor (DSP), Application-Specific Integrated Circuit (ASIC), Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. The general-purpose processor can be a microprocessor or any conventional processor, capable of implementing or executing the methods, steps, and logic block diagrams disclosed in the embodiments of this application.

[0069] The memory can be, but is not limited to, Random Access Memory (RAM), Read Only Memory (ROM), Programmable Read-Only Memory (PROM), Erasable Programmable Read-Only Memory (EPROM), Electrically Erasable Programmable Read-Only Memory (EEPROM), etc. The memory is used to store computer programs, and the processor can execute the computer programs accordingly after receiving execution instructions.

[0070] This application also provides a computer storage medium for storing the computer program used in the aforementioned computer device. The computer storage medium can be a readable storage medium, a non-volatile storage medium, or a volatile storage medium. For example, the computer storage medium may include, but is not limited to, various media capable of storing program code, such as a USB flash drive, a portable hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0071] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can also be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the flowcharts and block diagrams in the accompanying drawings show the architecture, functionality, and operation of possible implementations of apparatus, methods, and computer program products according to various embodiments of this application. 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 alternative implementations, the functions marked in the blocks may occur in a different order than those marked in the drawings. For example, two consecutive 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 diagram and / or flowchart, and combinations of blocks in the block diagram and / or flowchart, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0072] In addition, the functional modules or units in the various embodiments of this application can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part.

[0073] If the aforementioned functions are implemented as software functional modules and sold or used as independent products, they can be stored in a readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a smartphone, personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application.

[0074] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A cross-domain model communication method, characterized in that, The method is applied to a cross-domain model communication system, the system comprising: a user domain, a data domain, and a network gateway, and the method comprising: The user domain sends the request information for the model in the data domain to the gateway, and the gateway performs security audit on the request information. Upon successful security audit, the request information is sent to the data domain via the gateway, and the model generates response information based on the request information. The response information is sent to the gateway through the data domain, and the gateway performs security audit on the response information. Upon successful security audit, the response information is sent to the user domain via the gateway.

2. The cross-domain model communication method according to claim 1, characterized in that, The user domain includes a client and a front-end proxy. The step of sending request information for a model in the data domain to the gateway through the user domain includes: The client generates an interaction request for the model in the data domain and sends the interaction request to the front-end agent. The interaction request is encapsulated by the front-end proxy to generate request information with target identifier and interaction request content, and the request information is sent to the gateway.

3. The cross-domain model communication method according to claim 2, characterized in that, The data domain includes a caching layer and a back-end proxy. The step of sending the request information to the data domain through the gateway and generating response information based on the request information through the model includes: The request information is sent to the back-end proxy through the network gateway, and the back-end proxy parses the request information to obtain the interaction request content and the target identifier; The back-end proxy sends the interaction request content and the target identifier to the model, and the model generates response information based on the interaction request content. The model caches the response information into the cache layer based on the target identifier.

4. The cross-domain model communication method according to claim 3, characterized in that, The step of sending the response information to the gateway through the data field includes: The response information is split according to a preset splitting rule by the back-end proxy to obtain split response data; According to the splitting order, each splitting response data is sent to the gateway in sequence.

5. The cross-domain model communication method according to claim 4, characterized in that, The step of sending the response information to the user domain through the gateway includes: The gateway sends each of the split response data to the front-end agent sequentially based on the receiving order; The front-end agent sequentially sends each of the split response data to the client.

6. The cross-domain model communication method according to claim 1, characterized in that, The method further includes: In response to a query from the user domain, the data domain extracts the target data corresponding to the query from the cache layer of the data domain and sends the target data to the gateway. The target data is subjected to security audit through the network gateway, and when the security audit is passed, the target data is sent to the user domain through the network gateway.

7. The cross-domain model communication method according to claim 6, characterized in that, The method further includes: The target identifier is determined based on the receiving completion information or cache expiration instruction sent by the user domain in response to the data domain. Clear the cached data corresponding to the target identifier in the cache layer.

8. A cross-domain model communication device, characterized in that, The cross-domain model communication device includes: The first sending module is used to send request information for a model in the data domain to the network gateway through the user domain, and to perform security audit on the request information through the network gateway. The first review module is used to send the request information to the data domain through the network gateway when the security review is passed, and to generate response information based on the request information through the model. The second audit module is used to send the response information to the gateway through the data domain, and to perform security audit on the response information through the gateway. The second sending module is used to send the response information to the user domain through the network gateway when the security audit is passed.

9. A computer device, characterized in that, The computer device includes a processor and a memory, the memory storing a computer program, and the processor executing the computer program to implement the cross-domain model communication method according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when run on a processor, executes the cross-domain model communication method according to any one of claims 1-7.

Images