Service processing method in call based on 5G message and related equipment

Abstract

The invention provides a service processing method in a call based on a 5G message and related equipment, and the method comprises the steps: analyzing call content when an authorized user is in a call, determining a service scene corresponding to the current call, and extracting key information corresponding to the service scene from the call content; based on the service scene and the key information, calling an interface corresponding to an external service system, determining service data associated with the current call, and dynamically generating an interactive 5G message according to the service data; pushing the interactive 5G message to a call interface of the current call; and receiving an operation instruction generated when the user operates the interactive 5G message on the call interface, synchronizing the operation instruction to the external service system to execute service processing, and updating and presenting a processing result returned by the external service system in the call interface. Therefore, seamless connection of voice communication and service operation in the same session is realized, and the information interaction efficiency is remarkably improved.

Description

5G Messaging-Based In-Call Service Processing Methods and Related Equipment Technical Field

[0001] This application relates to the field of communication technology, and in particular to a method and related equipment for in-call service processing based on 5G messages. Background Technology

[0002] Currently, information exchange based on voice calls mainly relies on traditional voice communication technology. Both parties exchange information purely through voice, and key data such as addresses, amounts, and order numbers need to be manually recorded or confirmed later through other channels such as SMS or email. This results in low information transmission efficiency and a high risk of errors. Some systems attempt to assist with recording using simple voice-to-text functions, but they struggle to automatically recognize the business intent within the call. Furthermore, existing solutions disconnect call and data services. Users often need to interrupt the call and manually switch to other applications to query or confirm information, leading to cumbersome and time-consuming interaction processes. While some smart terminals support shortcut commands to access related services, users still need to actively trigger them, failing to achieve real-time analysis of call content and automatic service matching.

[0003] This indicates that there is a disconnect between the call process and data services. Voice communication and business operations cannot be seamlessly connected in the same session. Users must switch back and forth between the call and the application, making it difficult to complete real-time confirmation, query or transaction of information without interrupting the call, ultimately resulting in low information transmission efficiency. Summary of the Invention

[0004] The purpose of this application is to at least address one of the aforementioned technical defects, particularly the technical defect in the prior art where there is a disconnect between the call process and data services.

[0005] Firstly, this application provides a method for in-call service processing based on 5G messages. The method includes: when an authorized user is in a call, analyzing the collected call content to determine the service scenario corresponding to the current call, and extracting key information corresponding to the service scenario from the call content; based on the service scenario and key information, calling the interface of the corresponding external business system to determine the service data associated with the current call, and dynamically generating an interactive 5G message based on the service data, the interactive 5G message containing interactive elements for completing service operations; pushing the interactive 5G message to the call interface of the current call; receiving operation instructions generated by the user operating the interactive 5G message on the call interface, synchronizing the operation instructions to the external business system to execute service processing, and updating and presenting the processing results returned by the external business system on the call interface.

[0006] In one embodiment, the steps of analyzing the collected call content, determining the business scenario corresponding to the current call, and extracting key information corresponding to the business scenario from the call content include: performing preliminary screening of the collected call content through keyword matching to obtain candidate scenarios; using a deep learning-based semantic understanding model to perform semantic analysis on the call content, identify the user's true intent, and determine the business scenario corresponding to the current call from the candidate scenarios based on the identification results; and using a named entity recognition model, combined with entity tags associated with the business scenario, to extract key information from the call content.

[0007] In one embodiment, the step of calling the interface of the corresponding external business system based on the business scenario and key information to determine the business data associated with the current call includes: determining the external business system and its API interface corresponding to the business scenario from a pre-configured interface mapping relationship; when the external business system and its API interface pass security verification, constructing API request parameters based on key information; calling the API interface of the external business system using a standard protocol according to the API request parameters, obtaining the returned data from the external business system, and parsing the returned data to obtain the business data associated with the current call.

[0008] In one embodiment, the step of dynamically generating interactive 5G messages based on business data includes: obtaining an interactive message template matching the business scenario from a predefined template library, wherein the interactive message template contains one or more interactive elements; filling the interactive message template with business data through a dynamic rendering engine to generate interactive rich media message content; analyzing the interactive rich media message content and prioritizing the identified information regions; monitoring the current network status to obtain network quality parameters; performing layered compression processing on the interactive rich media message content according to the priority of each information region and the network quality parameters, and generating metadata containing compression parameters, priority tags, and network adaptation strategies; and encapsulating the layered compressed interactive rich media message content and the metadata together to form an interactive 5G message.

[0009] In one embodiment, the step of pushing an interactive 5G message to the call interface of the current call includes: detecting the 5G message support capability of the user's terminal device, determining the message push channel based on the detection result; and sending the interactive 5G message to the user's terminal device through the message push channel based on the network adaptation strategy contained in the interactive 5G message, so that the terminal device can hierarchically reconstruct the interactive rich media message content of the interactive 5G message according to the priority marker contained in the interactive 5G message, and present it to the call interface of the current call.

[0010] In one embodiment, the step of sending an interactive 5G message to a user's terminal device via a message push channel based on the network adaptation strategy contained in the interactive 5G message includes: splitting the interactive rich media message content of the interactive 5G message into multiple data blocks according to the priority marker contained in the interactive 5G message; and sending each data block to the user's terminal device via a message push channel according to the network adaptation strategy contained in the interactive 5G message.

[0011] In one embodiment, the steps of receiving operation instructions generated by a user interacting with interactive 5G messages on a call interface and synchronizing the operation instructions to an external business system to execute business processing include: capturing the user's interaction behavior with interactive 5G messages on the call interface through an event listening interface, and generating corresponding operation instructions based on the interaction behavior; synchronizing the operation instructions to an external business system through an encrypted channel so that the external business system can execute corresponding business processing according to the operation instructions.

[0012] Secondly, this application provides a service processing device for calls based on 5G messages. The device includes: a service scenario determination module, used to analyze the collected call content when an authorized user is making a call, determine the service scenario corresponding to the current call, and extract key information corresponding to the service scenario from the call content; an interactive 5G message generation module, used to call the interface of the corresponding external business system based on the service scenario and key information, determine the service data associated with the current call, and dynamically generate an interactive 5G message based on the service data, the interactive 5G message containing interactive elements for completing service operations; an interactive 5G message push module, used to push the interactive 5G message to the call interface of the current call; and a service processing result update module, used to receive the operation instructions generated by the user operating the interactive 5G message on the call interface, synchronize the operation instructions to the external business system to execute service processing, and update and present the processing results returned by the external business system on the call interface.

[0013] Thirdly, this application provides a storage medium storing computer-readable instructions, which, when executed by one or more processors, cause the one or more processors to perform the steps of any of the 5G message-based call service processing methods described in the above embodiments.

[0014] Fourthly, this application provides a computer device, including: one or more processors, and a memory; the memory stores computer-readable instructions, which, when executed by one or more processors, perform the steps of any of the 5G message-based call service processing methods described in the above embodiments.

[0015] As can be seen from the above technical solutions, the embodiments of this application have the following advantages: The call service processing method and related equipment based on 5G messages provided in this application, firstly, by analyzing the call content of authorized users in real time, determine the business scenario corresponding to the current call and extract key information, thereby accurately identifying the actual business needs generated by the user during the call, avoiding the tedious process of the user having to manually repeat the description or switch applications to re-enter the information afterward; secondly, based on the identified business scenario and key information, call the interface of the corresponding external business system to obtain the business data associated with the current call, and dynamically generate an interactive 5G message containing interactive elements based on the business data, introducing the data that was originally isolated from the external business system into the call process in real time, and transforming static information into an operable interactive interface, so that the user can directly obtain the business data related to the call content without interrupting the call, breaking the disconnect between call and data services. This application breaks down information barriers between services; furthermore, it pushes the generated interactive 5G message to the current call interface, allowing users to see and operate the message intuitively while maintaining call continuity; it enables parallel operation of voice communication and visual business operations, eliminating the operational gaps that users need to repeatedly switch between calls and applications, significantly improving the continuity and convenience of interaction; finally, it receives the operation instructions generated by the user's operation of the interactive 5G message on the call interface, synchronizes the instructions to the external business system for business processing, and updates the processing results returned by the external business system in the call interface, allowing users to complete the entire process from business initiation to business execution to result confirmation during the call, without the need for secondary confirmation or repeated operations afterward, truly achieving real-time information confirmation, query, or transaction without interrupting the call, fundamentally solving the technical problem of low information transmission efficiency. In summary, this application, by introducing a closed-loop feedback mechanism for business data acquisition based on real-time semantic analysis, interactive message generation and push, and user operation instructions during the call, achieves seamless connection between voice communication and business operations in the same session, significantly improving information interaction efficiency. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 is a flowchart illustrating the in-call service processing method based on 5G messages provided in an embodiment of this application; Figure 2 is a structural diagram illustrating the in-call service processing device based on 5G messages provided in an embodiment of this application; Figure 3 is a schematic diagram illustrating the internal structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0018] The technical solutions of 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. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0019] This application provides a method for in-call service processing based on 5G messaging. The following embodiments illustrate this method using a computer device as an example. It is understood that the computer device can be any device with data processing capabilities, including but not limited to a single server and server clusters. As shown in Figure 1, the method may include the following steps: S101: When an authorized user is making a call, the collected call content is analyzed to determine the service scenario corresponding to the current call, and key information corresponding to the service scenario is extracted from the call content.

[0020] In this application, an authorized user refers to a user who has pre-authenticated their identity and actively enabled the call content processing function. This authorization process typically involves the user completing identity verification through their terminal device and agreeing to relevant privacy terms, ensuring that their call content is analyzed only with their knowledge and permission. Call content refers to audio data generated by the user during a call, after being collected and pre-processed, as well as text data streams obtained through further conversion using speech recognition technology. Business scenario refers to the specific service type corresponding to the user's intent identified through semantic analysis of the call content, such as order inquiry, address modification, payment confirmation, and logistics tracking. Key information refers to the core data elements necessary to complete business operations in a specific business scenario, such as order number, amount, account, address, and phone number.

[0021] In the implementation process, when an authorized user begins a call, the system first checks whether the user has enabled call content processing. If so, it collects the voice signal during the call in real time. The collected voice signal undergoes preprocessing such as noise suppression and echo cancellation before being fed into the speech recognition engine to convert it into a text data stream. To ensure accuracy, an end-to-end speech recognition model based on deep learning is used. This model has been trained on a large amount of multilingual and accent data and can adapt to the pronunciation habits of different users. Simultaneously with speech recognition, the speaker's role can be distinguished, for example, by identifying whether it is the user or the other party through voiceprint features, so as to more accurately understand the context in subsequent analysis. After obtaining the text data stream, it undergoes text preprocessing, including spell correction, word segmentation, stop word removal, and conversion of numbers, dates, etc., into a unified format to reduce ambiguity in subsequent analysis.

[0022] In one implementation, a joint deep learning model is used to identify business scenarios and extract key information. Specifically, a large amount of pre-annotated speech data is collected, with each segment labeled with its corresponding business scenario and the start and category of key information. A multi-task deep learning model, such as a BERT-based joint model, is trained using this data. This model simultaneously outputs business scenario classification results and sequence labeling results for key information. When an authorized user makes a call, speech is captured in real-time and converted into a text data stream. After the same text preprocessing, the preprocessed text is input into the trained joint model. The model first extracts deep semantic features of the text through a multi-layer Transformer encoder, then outputs the probability distribution of business scenarios through a classification layer, selecting the scenario with the highest probability as the business scenario corresponding to the current call. Simultaneously, the model performs label prediction for each word or character through a sequence labeling layer, for example, using the BIO labeling system to mark the start and end positions of key information and identify its category. Taking an order query scenario as an example, the model might label "order number" as the key information type and the following string of numbers "12345" as the corresponding value. The model outputs structured key information, eliminating the need to rely on predefined extraction templates. To further improve accuracy, the model output can be post-processed, such as validating the format of extracted numbers. If obvious errors are found, a review process is triggered. This end-to-end joint model can fully utilize contextual information, adapt to diverse modes of expression, and is particularly suitable for handling conversational and disjointed dialogue.

[0023] After identifying the business scenario and extracting key information, the results are associated with and stored using the current call's session identifier for subsequent steps. The entire analysis process is conducted with user authorization, strictly adhering to privacy regulations and processing only information explicitly authorized by the user. Furthermore, the deep learning model can be retrained according to business needs to adapt to emerging business scenarios and expressions, ensuring the accuracy and timeliness of identification and extraction.

[0024] It's worth noting that by analyzing authorized users' call content in real time, the system can accurately identify users' business intentions during the call and extract key information needed to complete the task. This process avoids the tedious steps of users manually inputting or switching applications to repeat their requirements afterward, significantly improving the efficiency and accuracy of information acquisition. Simultaneously, by converting voice interaction into structured business data and scenario tags, a reliable data foundation is provided for subsequent automatic calls to external business systems, ensuring seamless business operations. Ultimately, users can trigger subsequent service processes without interrupting the call, effectively solving the information gap between call and data services and improving overall interaction efficiency.

[0025] S102: Based on the business scenario and key information, call the interface of the corresponding external business system to determine the business data associated with the current call, and dynamically generate an interactive 5G message based on the business data. The interactive 5G message contains interactive elements for completing business operations.

[0026] In this application, an external business system refers to a third-party system that is independent of the current call system and can provide specific business services, such as an e-commerce platform order system, a bank payment system, or a logistics company query system. An interface refers to a standardized data access channel provided by the external business system, typically manifested as an Application Programming Interface (API) conforming to protocols such as RESTful or SOAP. Business data refers to information directly related to the user's current needs obtained through interface calls, such as order status, logistics tracking, account balance, and payment confirmation page. Interactive 5G messaging refers to rich media messages conforming to 5G messaging standards, which not only include multimedia content such as text, images, and videos, but also embed an interactive interface. Interactive elements refer to user-operable components included in interactive 5G messages, such as fillable electronic forms, clickable confirmation buttons, and selectable option lists. These elements allow users to complete business operations directly on the message interface without needing to navigate to other applications.

[0027] After identifying the business scenario and extracting key information, a service registration and discovery mechanism can be used to dynamically invoke external business system interfaces. A service registry is maintained, where each external business system registers its available interface service information in real time, including the interface address, supported business scenarios, input parameter format, output data format, and current service status. Once a business scenario is identified and key information is extracted, instead of relying on a static pre-configured mapping table, a query request is sent to the service registry to obtain a list of currently available interface services that support the business scenario. Based on a preset strategy, the optimal interface is selected from the returned available interfaces, such as prioritizing the service instance with the shortest response time or lightest load. After determining the target interface, the real-time access address and required dynamic authentication credentials, such as token-based dynamic access credentials or one-time keys, are obtained from the registry. Subsequently, the extracted key information is assembled into request parameters according to the interface's required format, and the call is initiated. During the call, the interface's response time and success rate are monitored in real time. If the current interface is slow to respond or the call fails, it can automatically switch to other available interfaces returned by the registry for retry. After processing the request, the external business system returns response data. This response is received and parsed to extract the business data associated with the current call. This dynamic invocation method adapts to changes in the business system; for example, when an interface service is upgraded or migrated, only the information in the registry center needs to be updated, without modifying the local configuration. Simultaneously, real-time monitoring and automatic switching mechanisms ensure high availability of the interface calls, guaranteeing that users can reliably obtain the necessary business data during the call. After obtaining the business data, the message generation phase begins.

[0028] In one implementation, a component library can be maintained, containing various independently usable interactive components, such as information display cards, confirmation buttons, selection lists, and form input boxes. Each component has predefined styles and capabilities, and its appearance and behavior can be dynamically adjusted through configuration parameters. Upon acquiring the business scenario and key information, the system makes real-time decisions on which components to assemble and how to combine them, based on the scenario type and the structure of the business data. For example, in a logistics query scenario, it might decide to assemble a card component displaying the logistics trajectory, a button component for contacting the courier, and a selection component for confirmation of receipt. The configuration engine injects business data into the corresponding fields of each component; for example, logistics node information is filled into the trajectory card, and the courier's phone number is filled into the dialing link of the button. Subsequently, the layout engine combines the selected components into a complete rich media interface according to a predefined or dynamically calculated arrangement, generating a corresponding 5G message description file. This dynamic assembly method can adapt to more complex and varied business needs, especially suitable for scenarios requiring personalized interaction flows. Regardless of whether template filling or dynamic assembly is used, the generated interactive 5G messages contain clear interactive semantics, allowing users to intuitively understand how to operate without additional learning. After the message is generated, a lightweight compatibility check is performed to ensure that its main interactive elements can be displayed and responded normally on mainstream terminals.

[0029] In another implementation, a multi-step interactive message generation strategy is adopted for complex business scenarios. For example, in a scenario requiring user confirmation of multiple pieces of information, an overview interactive message is first generated, displaying all key information to be confirmed and including an edit button. After the user clicks the edit button, a second-level interactive message is dynamically generated according to a preset interaction flow, such as an editing interface containing a detailed form. This multi-level interaction design allows complex business processes to be broken down into multiple simple steps, requiring the user to process only a small amount of information each time, reducing the operational burden. By maintaining a dialogue state machine to track the current interaction level, it is ensured that each operation correctly leads to the next step. All generated interactive messages are associated with the same call session, allowing the user to complete the entire business process seamlessly during the call without interruption.

[0030] It's important to note that by accurately invoking the interfaces of corresponding external business systems based on identified business scenarios and key information, real-time access to genuine business data closely related to the user's current needs can be obtained, avoiding the lag and inaccuracy that may exist with pre-set static content. After acquiring the data, 5G messages containing interactive elements are dynamically generated, transforming business information that was originally isolated from external systems into an interactive interface that users can directly operate. This process ensures that the message content pushed to the user is not only authentic and effective but also highly matched to the specific needs arising during the call. Users can intuitively view information and complete subsequent operations directly through the interactive elements in the message without leaving the call interface. This approach upgrades the traditional one-way information display to two-way business interaction, significantly improving the efficiency and accuracy of information transmission and providing a technical foundation for users to seamlessly connect to data services without interrupting their calls.

[0031] S103: Push interactive 5G messages to the call interface of the current call.

[0032] The call interface refers to the terminal screen display area that the user sees during a call. This area typically includes basic elements such as the other party's avatar, call duration, and call control buttons. In this application, this area is also used to display interactive 5G messages pushed to the user, allowing the user to view and operate the message while keeping the call going.

[0033] In one implementation, after generating an interactive 5G message, the session identifier of the current call is first obtained. This identifier is assigned during call establishment and used to uniquely identify the current call. The interactive 5G message is then associated with and stored with the session identifier to ensure that the message is accurately attributed to the current call. Subsequently, the 5G message push interface provided by the operator platform is invoked to send the interactive 5G message to the user terminal in the form of a data packet. During the sending process, the terminal's reception status is monitored in real time. If a delivery confirmation is received from the terminal, it indicates that the message has been successfully pushed to the call interface; if no confirmation is received within a preset time, a retry strategy is automatically executed to attempt to resend the message. To ensure timely message delivery, the sending strategy is dynamically adjusted according to network conditions, such as appropriately reducing the sending priority while ensuring eventual delivery during network congestion. When the user terminal successfully receives the interactive 5G message, the terminal operating system will directly present it in the current call interface according to the message type, usually in the form of a half-screen card or a floating window, avoiding complete obscuring of the call content. The user can see the message on the call interface and perform operations directly. During the operation, the call remains connected, and the user does not need to hang up the phone or switch to other applications.

[0034] In another implementation, the timing of push notifications is dynamically adjusted based on the call status. The call process is monitored in real time, for example, by acquiring status changes such as call start, call hold, call transfer, and call about to end through signaling monitoring. When the call is detected to be about to end, core business summary information from interactive 5G messages is prioritized and pushed to the call interface, ensuring that the user receives the most important results before hanging up. Non-urgent detailed information can be sent as a regular message after the call ends. This dynamic push strategy based on call status maximizes the chances of users receiving critical business information before the call ends, avoiding information loss due to call interruptions. Simultaneously, the push priority of subsequent messages is intelligently adjusted based on user feedback on the call interface; for example, if the user quickly closes a certain type of message, the push priority of that type of message will be reduced.

[0035] It's worth noting that pushing interactive 5G messages directly to the current call interface allows users to view relevant business information in real time while maintaining the continuity of the voice call. This avoids the cumbersome process of hanging up the phone or switching to other applications to access information, ensuring the immediate accessibility of business data. With messages displayed on the call interface, users can intuitively understand the business progress without interrupting the current conversation and can refer to the message content at any time during the call for subsequent communication. This design effectively solves the problem of the separation of information and services in time and space during a call, significantly improving the efficiency of information delivery and the continuity of the user experience.

[0036] S104: Receive operation instructions generated by the user's operation of interactive 5G messages on the call interface, synchronize the operation instructions to the external business system to execute business processing, and update and present the processing results returned by the external business system on the call interface.

[0037] In this application, an operation instruction refers to a digital instruction generated by a user interacting with interactive 5G messages displayed on the call interface, such as a confirmation instruction generated by clicking a confirmation button, a submission instruction generated after filling out a form and submitting it, or a selection instruction generated after selecting an option. Business processing refers to the specific business logic executed by an external business system based on the received operation instructions, such as deducting payment, changing an address, generating an order, or updating the status. The processing result refers to the status information returned by the external business system after completing the business processing, such as success confirmation, reason for failure, information to be supplemented, and updated data.

[0038] In one implementation, a complete reliability assurance system is established for the transmission and processing of operation instructions. When synchronizing operation instructions to external business systems, a unique instruction identifier is generated for each instruction and associated with the call session identifier. The sending time, target system, instruction content, and current status (e.g., sent and awaiting response) of the instruction are recorded. A timeout monitoring timer is started; if no processing result response is received from the external business system within a preset time, a retry mechanism is automatically triggered, resending the instruction with an exponential backoff strategy increasing the retry interval until success or the maximum number of retries is reached. During the retry process, it is ensured that the same instruction is not executed repeatedly in the external business system, for example, through idempotency control using the instruction identifier. If the maximum number of retries is reached and the task still fails, the instruction status is marked as failed, and a failure notification is pushed to the user in the call interface, prompting the user to retry later or contact customer service. For instructions that have been successfully received but require subsequent feedback on the processing result, a result callback listening mechanism is established to ensure that the external business system can accurately callback the system interface after completing processing. Detailed logs are recorded for the entire process of sending, retrying, succeeding, and failing all commands. These logs include key information such as command identifier, session identifier, timestamp, operation type, and return result. This structured log data can be used for subsequent auditing, tracing, and process optimization. Users do not need to concern themselves with the backend reliability mechanisms throughout the process. This design ensures that user operations are reliably executed and the final results are accurately presented.

[0039] In another implementation, various processing mechanisms are flexibly adopted based on the specific characteristics of the business corresponding to the user's operation command to ensure the reliability and security of business execution. For business scenarios requiring long-term processing by external business systems, such as manual review or interbank clearing, an asynchronous processing mechanism is used. After capturing the user's operation command and completing preliminary verification, a confirmation message indicating that the command has been received is immediately sent to the user. Simultaneously, the message status is updated to "processing" in the call interface, and the command is associated with the current call session identifier and placed in the pending queue. The command is forwarded to the external business system through a reliable message queue mechanism, and the system waits for the processing result through polling or callback. After the external business system completes processing, the system receives the result and updates the message presentation in the call interface. The user does not need to hang up the phone to wait throughout the process. For complex business scenarios involving multiple external business systems, such as order processing that simultaneously requires deducting inventory, deducting payment, and updating logistics status, a distributed transaction coordination mechanism is used. A globally unique distributed transaction identifier is generated for this business request, and the interfaces of each external system are called sequentially or in parallel according to the preset transaction process to execute sub-operations, while recording the status of each sub-operation. If all sub-operations execute successfully, a transaction commit confirmation is sent to each system, and the success result is updated in the call interface. If any sub-operation fails, a compensation process is immediately initiated, calling the rollback interface of the successful systems to ensure eventual data consistency, and the reason for the failure is fed back to the user. For high-risk and sensitive business scenarios, such as large-amount transfers or modifications to critical account information, risk control and secondary confirmation mechanisms are integrated. First, the risk control model is called to perform a real-time assessment of the operation. If the risk assessment passes, a secondary confirmation request is generated and pushed to the call interface in the form of a new interactive message, requiring the user to confirm the operation intention again through SMS verification code, facial recognition, etc. Only after successful verification will the system synchronize the original operation instruction to the external business system for execution and update the processing result. If the risk assessment fails or the secondary verification fails, the operation is directly rejected and the reason is displayed to the user. These processing mechanisms can be used individually or in combination according to the needs of the business scenario to ensure that user operations are processed safely, reliably, and efficiently during the call.

[0040] It's worth noting that by receiving user-generated interactive 5G messages within the call interface in real time, synchronizing these commands to external business systems for processing, and simultaneously updating the processing results on the call interface, a complete closed loop is achieved from user operation to business execution and result feedback. This process ensures that users can complete business operations without hanging up or switching to other applications, avoiding information delays and redundant communication caused by operational gaps in traditional methods. The instant updates of processing results allow users to know the business execution status immediately, eliminating information uncertainty and improving the consistency and satisfaction of the user experience. Simultaneously, synchronizing commands to external business systems for direct execution ensures the real-time nature and accuracy of operations, avoiding errors that might be introduced by manual relaying or subsequent processing. Ultimately, this design effectively bridges the gap between the call process and data services, enabling seamless integration of voice communication and business operations within the same session.

[0041] In the above embodiments, firstly, by analyzing the call content of authorized users in real time, the business scenario corresponding to the current call is determined and key information is extracted, thereby accurately identifying the actual business needs generated by the user during the call, avoiding the tedious process of the user having to manually repeat the description or switch applications to re-enter the information afterward; secondly, based on the identified business scenario and key information, the interface of the corresponding external business system is called to obtain business data associated with the current call, and an interactive 5G message containing interactive elements is dynamically generated based on the business data, bringing the data that was originally isolated from the external business system into the call process in real time, and transforming static information into an operable interactive interface, so that the user can directly obtain business data related to the call content without interrupting the call, breaking down the information barrier between call and data services; furthermore, the generated interactive 5G message is... The message is pushed to the current call interface, allowing users to see and interact with it intuitively while maintaining call continuity. This enables parallel voice communication and visual business operations, eliminating the operational gaps that require users to repeatedly switch between calls and applications, significantly improving the continuity and convenience of interaction. Finally, the system receives operation commands generated by the user's interaction with the interactive 5G message on the call interface and synchronizes these commands to the external business system for execution. Simultaneously, the processing results returned by the external business system are updated and presented on the call interface. This allows users to complete the entire process from business initiation to execution and result confirmation during the call, without the need for subsequent confirmation or repetitive operations. It truly achieves real-time information confirmation, querying, or transactions without interrupting the call, fundamentally solving the technical problem of low information transmission efficiency. In summary, this application, by introducing a closed-loop feedback mechanism based on real-time semantic analysis for business data acquisition, interactive message generation and push, and user operation commands during the call, achieves seamless integration of voice communication and business operations within the same session, significantly improving information interaction efficiency.

[0042] In one embodiment, the steps of analyzing the collected call content, determining the business scenario corresponding to the current call, and extracting key information corresponding to the business scenario from the call content include: performing preliminary screening of the collected call content through keyword matching to obtain candidate scenarios; using a deep learning-based semantic understanding model to perform semantic analysis on the call content, identify the user's true intent, and determine the business scenario corresponding to the current call from the candidate scenarios based on the identification results; and using a named entity recognition model, combined with entity tags associated with the business scenario, to extract key information from the call content.

[0043] Candidate scenarios refer to one or more business categories that may match the user's intent, initially selected through keyword matching of the call content, such as order inquiries, logistics tracking, and payment confirmation. Semantic understanding models refer to deep learning-based natural language processing models, such as pre-trained models like BERT or GPT. These models can perform deep semantic analysis of text and understand the user's true intent in conjunction with the context. True intent refers to the actual business needs expressed by the user through voice. For example, if a user says, "I want to check where the red dress I bought last week is," their true intent is to check the logistics status of a specific order. Named entity recognition models are deep learning models used to identify and extract specific types of entities from text. These models can identify information fragments with specific meanings, such as names, places, dates, order numbers, and amounts. Entity tags are information category identifiers preset by the named entity recognition model for different business scenarios, such as order numbers, amounts, and shipping addresses.

[0044] In one implementation, after receiving the call content from an authorized user and completing text preprocessing, a scene recognition process is initiated. First, a pre-built keyword library is loaded, covering core vocabulary for various business scenarios. For example, the order query scenario includes keywords such as "order," "check," and "order number," while the payment confirmation scenario includes keywords such as "payment," "payment," and "deduction." The preprocessed text is then segmented into words, and the segmentation results are compared one by one with the keyword library. The number of keyword hits for each business scenario is counted, and a matching score is calculated for each scenario based on the number of hits and keyword weights. Scenarios with scores exceeding a preset threshold are designated as candidate scenarios. For example, if the text contains "order" and "check," the order query scenario has a higher score and is included in the candidate list. This initial screening process can quickly narrow down the scope when processing large-scale corpora, avoiding the need for subsequent complex deep learning models to analyze all possible scenarios one by one, significantly improving processing efficiency.

[0045] After initial screening to identify candidate scenarios, the process moves to deep semantic analysis. The pre-processed complete text is input into a pre-trained semantic understanding model. This model, based on pre-trained architectures like BERT and fine-tuned with extensive labeled call data, accurately understands colloquial expressions and complex sentence structures. The model extracts deep semantic features from the text using a multi-layer Transformer encoder and analyzes the user's true intent in conjunction with the dialogue history. For example, when a user says, "I want to check where the red dress I bought last week is," the model not only recognizes the action of "checking," but also understands, based on the description of the red dress, that the user is inquiring about the logistics status of a specific item rather than making a general inquiry. The model outputs a confidence score for each candidate scenario and selects the scenario with the highest confidence score as the business scenario corresponding to the current call. If the confidence scores of all candidate scenarios are below a preset threshold, a clarification process may be triggered, such as generating an inquiry message to allow the user to further clarify their needs.

[0046] After determining the business scenario, the key information extraction stage begins. Based on the identified business scenario type, a set of entity tags associated with that scenario is dynamically loaded. For example, for an order query scenario, entity tags might include order number, product name, and order time; for a payment confirmation scenario, entity tags might include amount, payee, and payment method. The preprocessed text is then input into a named entity recognition model fine-tuned for that business scenario. This model uses a sequence labeling algorithm to predict the tag for each word or character in the text, marking the text fragments belonging to each entity tag and their positions. For example, for the text "I want to query order number 12345", the model might tag 12345 as the order number entity. The model outputs all identified entities and their corresponding tag categories, organizing these entities into structured key information according to tag categories, such as generating key-value pairs for order number 12345. To ensure the accuracy of the extraction, format and logical verification can be performed on the extraction results. For example, checking whether the order number conforms to the length and verification rules. If incomplete or incorrectly formatted information is found, the system can record the anomaly and trigger supplementary inquiries or manual review processes.

[0047] It's important to note that initial screening of call content through keyword matching quickly narrows down potential business scenarios, reducing the amount of data that subsequent complex models need to process, thereby improving overall response efficiency. Based on the selected candidate scenarios, a deep learning-based semantic understanding model is used for in-depth analysis. This accurately understands users' colloquial and disjointed expressions, precisely identifying their true intentions and avoiding potential misjudgments from keyword matching. After determining the business scenario, a named entity recognition model combined with customized entity tags for that scenario is used to extract key information from the call content. This ensures that the obtained data is closely related to specific business needs and is presented in a structured form for easy use by subsequent systems. This coarse-to-fine, progressive processing approach guarantees both the accuracy of intent recognition and the targeted extraction of key information, providing a reliable data foundation for seamless integration with external business systems.

[0048] In one embodiment, the step of calling the interface of the corresponding external business system based on the business scenario and key information to determine the business data associated with the current call includes: determining the external business system and its API interface corresponding to the business scenario from a pre-configured interface mapping relationship; when the external business system and its API interface pass security verification, constructing API request parameters based on key information; calling the API interface of the external business system using a standard protocol according to the API request parameters, obtaining the returned data from the external business system, and parsing the returned data to obtain the business data associated with the current call.

[0049] The interface mapping relationship refers to a pre-established and stored table corresponding to business scenarios and external business system application programming interfaces (APIs). This table records which external business system's API should be called for each business scenario. An API is a standardized data access channel exposed by an external business system, typically following RESTful or SOAP protocols. Requests can be sent and responses received through this channel. Security verification refers to the process of authenticating and checking the permissions of the external business system interface to be called, ensuring that only authorized systems can access the interface. Request parameters are a data set assembled according to the interface specification based on key information, usually organized in JSON or XML format, containing all necessary fields required for interface execution. Standard protocols refer to standardized data transmission protocols widely used in internet communication, such as HTTPS, which ensures the security and reliability of data during transmission. Returned data refers to the raw response information returned by the external business system after processing the request, typically including status codes and data bodies.

[0050] In one implementation, after determining the business scenario corresponding to the current call and extracting key information, the pre-configured interface mapping relationship is first queried based on the business scenario. This mapping relationship is stored in the form of a configuration file or database table, and each record contains information such as the business scenario identifier, the corresponding external business system name, the API interface address, the request method, and the authentication method. For example, the order query scenario corresponds to the query interface of the order system, and the payment confirmation scenario corresponds to the transaction confirmation interface of the payment system. The unique target interface record is located by searching the mapping table through the scenario identifier. After obtaining the interface information, the security verification stage begins. The authentication credentials required to access the interface are obtained from the unified credential management service, such as OAuth 2.0 access tokens, API keys, or digital certificates. The validity of the credentials is checked, and if the credentials have expired, they are automatically refreshed or reapplied for. After successful verification, the extracted key information is assembled into request parameters according to the format defined in the interface document. For example, for the order query interface, the value of the order number field is put into the corresponding key of the JSON object, and necessary public parameters such as timestamp and request identifier are added. After the parameters are assembled, a request is sent to the API interface address of the external business system using the HTTPS protocol. The request method is usually GET or POST, and the authentication credentials are carried in the request header. Set a reasonable timeout, such as 3 seconds, to avoid slow interface response affecting the call experience. After receiving the request, the external business system executes the corresponding business logic, such as querying order information in the database, and then returns the result in a standard format. Upon receiving the returned data, first check the HTTP status code. If the status code indicates success, parse the returned data body to extract the business data. The parsing process is based on the data structure returned by the interface, such as retrieving fields like order status and tracking number from a JSON object. The parsed business data is associated with the current call's session identifier and stored, ready for subsequent generation of interactive 5G messages. If a network exception or error code occurs during the call, retry according to a preset strategy. The number of retries is usually no more than 3. If it still fails, record the exception and trigger an alarm.

[0051] It's important to note that by using pre-configured interface mapping relationships to determine the external business systems and their API interfaces corresponding to the business scenario, the accuracy and relevance of calls are ensured, avoiding data retrieval failures due to system mismatches. After security verification, request parameters are constructed based on key information, guaranteeing that only authorized and legitimate requests can access external systems, while also ensuring that the information carried in the request is complete, accurate, and conforms to interface specifications. Using standard protocols to call the interface and parsing the returned data ensures the security and reliability of data transmission, and transforms the raw response into structured business data for direct subsequent use. This series of designs enables the system to acquire real-time business data matching user needs during the call, providing an accurate and reliable information foundation for seamless data service integration.

[0052] In one embodiment, the step of dynamically generating interactive 5G messages based on business data includes: obtaining an interactive message template matching the business scenario from a predefined template library, wherein the interactive message template contains one or more interactive elements; filling the interactive message template with business data through a dynamic rendering engine to generate interactive rich media message content; analyzing the interactive rich media message content and prioritizing the identified information regions; monitoring the current network status to obtain network quality parameters; performing layered compression processing on the interactive rich media message content according to the priority of each information region and the network quality parameters, and generating metadata containing compression parameters, priority tags, and network adaptation strategies; and encapsulating the layered compressed interactive rich media message content and the metadata together to form an interactive 5G message.

[0053] The template library refers to a pre-built and stored collection of interactive message styles designed for different business scenarios. Each template defines the message's layout structure, visual style, and the position and type of interactive elements. The dynamic rendering engine is a program component that generates a visual interface in real time based on business data and template definitions. It can fill structured business data into specified positions in the template and generate rich media content compliant with 5G messaging standards. Interactive rich media message content refers to a complete message interface description containing specific business data and interactive elements after rendering. Information areas refer to semantically independent components within the message content, such as text blocks, image areas, button groups, and form fields. Priority classification is the process of hierarchically labeling different areas according to the importance of the information. For example, core operation buttons and key information are marked as high priority, while background decorative elements are marked as low priority. Network quality parameters are quantitative indicators reflecting the current network connectivity status, including available bandwidth, round-trip latency, and packet loss rate. Layered compression processing is a technique that uses different compression ratios for different information areas based on priority and network conditions. High-priority areas use lossless or near-lossless compression to ensure quality, while low-priority areas use high compression ratios to reduce data volume. Compression parameters refer to configuration data that records information such as the type of compression algorithm, compression ratio, and quality factor used during the compression process. Priority tags are labels that identify the importance of each information area, used to guide subsequent transmission and reconstruction processes. Network adaptation strategies are a set of rules that dynamically adjust transmission behavior based on network quality, such as prioritizing the transmission of high-priority data when network bandwidth decreases. Metadata refers to descriptive information containing compression parameters, priority tags, and network adaptation strategies; it is encapsulated along with the message content for use by the receiving end during parsing and presentation.

[0054] In one implementation, the message generation stage begins after the business data is acquired. First, a predefined template library is queried based on the previously determined business scenario. This library stores interactive message templates designed for various business scenarios; for example, an order query scenario corresponds to a template including an order status display card and a logistics details button, while a payment confirmation scenario corresponds to a template including an amount display area and a payment confirmation button. The template file matching the current business scenario is retrieved from the template library. This template, described in JSON or XML format, defines the overall layout of the message, the position and size of each area, and the types of embedded interactive elements. After retrieving the template, a dynamic rendering engine is started to populate the corresponding positions in the template with the business data. The rendering engine parses the template structure, identifies placeholders that need data filling, such as order number placeholders and amount placeholders, and then replaces them with the actual values ​​from the business data. Simultaneously, the rendering engine dynamically configures the behavior of interactive elements based on the attributes of the business data; for example, it sets the style of the confirmation button based on the payment amount and determines whether to display the edit button based on the order status. After rendering is complete, a complete interactive rich media message is generated. This content is described in the 5G message standard format and includes all visual elements and interactive logic.

[0055] After generating the message content, the content optimization phase begins. Deep analysis is performed on the generated interactive rich media message content to identify various information regions. The analysis process utilizes image processing and layout parsing techniques, such as detecting text blocks, image areas, button groups, and form fields, and performing semantic understanding on each region to determine its information type and importance. Based on the analysis results, the identified information regions are prioritized; for example, core operation buttons and key data display areas are marked as high priority, while background decorations and supplementary explanatory text are marked as medium or low priority. After the prioritization, priority markers are appended to the metadata of each region.

[0056] Simultaneously, the current network status is monitored, and real-time network quality parameters, including available bandwidth, round-trip latency in milliseconds, and packet loss percentage, are obtained through communication interfaces with terminal devices or network protocol stacks. These parameters are recorded for subsequent compression decisions. After obtaining the priority division results and network quality parameters, hierarchical compression processing is initiated. The compression process dynamically selects a compression strategy based on the priority labels of each information area and the current network conditions: for high-priority areas, such as the payment confirmation button and amount display area, lossless or near-lossless compression algorithms are used to ensure that the images and text in these areas remain clearly legible after transmission; for medium-priority areas, such as product thumbnails, moderate compression is used to reduce the data volume while ensuring basic recognizability; for low-priority areas, such as background textures or decorative elements, high compression ratios or even partial removal are used to significantly reduce the data volume. When network quality parameters indicate low bandwidth, the compression ratio is further adjusted, for example, by appropriately increasing the compression ratio for all areas, but always ensuring that the quality of high-priority areas is prioritized. After compression, a complete set of metadata is generated. This metadata includes compression parameters used in each region, such as algorithm type and compression ratio, priority markers for each region, and network adaptation strategies based on network conditions, such as prioritizing the transmission of high-definition versions after network recovery. Finally, the layered compressed interactive rich media message content and the generated metadata are encapsulated together into a standard 5G message data packet. This data packet conforms to the GSMA RCS specification and includes a message header, message body, and metadata, ready for subsequent push processes.

[0057] In another implementation, redundant information is identified and removed before compression. Visual analysis is performed on image regions within the interactive rich media message content to identify repetitive content, solid color areas, and regular patterns. For example, for large monochrome backgrounds, instead of directly compressing pixel data, color values ​​and area ranges are recorded and described as solid color fill instructions in the metadata; for periodic texture patterns, their pattern characteristics are analyzed and parameterized descriptions are stored instead of complete image data. This redundancy removal significantly reduces the amount of data that needs to be compressed and transmitted, especially suitable for rich media messages containing a large amount of background or repetitive elements. After redundancy removal, the remaining core content is processed according to a layered compression strategy, and the removal instructions are also encapsulated as part of the metadata. The receiving end can then reconstruct the background and texture based on these instructions during reconstruction.

[0058] In another implementation, multi-version compression is performed on dynamically generated interactive rich media message content. Based on different preset values ​​for network quality parameters, multiple compressed versions are pre-generated for the message content, such as a high-quality version for high-speed networks, a medium-quality version for ordinary networks, and a basic version for weak network environments. Each version contains complete metadata describing its compression parameters and priority flags. When the current network status is detected, the most suitable compressed version is directly selected for encapsulation and push, eliminating the need for real-time compression calculations and further reducing processing latency. When network status changes, versions can be switched or content quality can be gradually upgraded through supplementary transmission mechanisms.

[0059] It's worth noting that by retrieving interactive message templates from a template library that match the business scenario and populating them with business data, personalized message content highly tailored to user needs can be quickly generated, ensuring that the information users see is directly relevant and contains actionable elements. Information area analysis and priority division of the generated rich media content can identify core operational areas and key display information, laying the foundation for subsequent differentiated processing. Combined with real-time monitored network quality parameters, the hierarchical compression strategy is dynamically adjusted according to the priority of each information area. This effectively reduces the data volume of secondary content while ensuring the transmission quality of core information, avoiding message transmission delays or failures due to network fluctuations. Compression parameters, priority markers, and network adaptation strategies are encapsulated as metadata and sent along with the message, enabling the receiving end to accurately parse and intelligently reconstruct the content based on current conditions. This ensures that users can obtain a clear and usable interactive interface in a timely manner under different network environments, thereby improving overall transmission efficiency and user experience.

[0060] In one embodiment, the step of pushing an interactive 5G message to the call interface of the current call includes: detecting the 5G message support capability of the user's terminal device, determining the message push channel based on the detection result; and sending the interactive 5G message to the user's terminal device through the message push channel based on the network adaptation strategy contained in the interactive 5G message, so that the terminal device can hierarchically reconstruct the interactive rich media message content of the interactive 5G message according to the priority marker contained in the interactive 5G message, and present it to the call interface of the current call.

[0061] Among these, the 5G messaging support capability of the terminal device refers to whether the user's mobile terminal has the ability to receive and parse 5G messages, which depends on the terminal model, operating system version, and operator network support. The message push channel refers to the communication path used to deliver interactive 5G messages to the user terminal, including different methods such as 5G messaging channels, ordinary SMS channels, or in-app notifications. Hierarchical reconstruction refers to the process by which the terminal device parses and presents message content in stages according to the received priority markers, prioritizing the restoration of high-priority areas to ensure that users can quickly obtain core information.

[0062] In one implementation, the push notification process is initiated after the interactive 5G message is generated. First, a probe signaling is sent to the user terminal, or a locally maintained terminal capability database is queried to determine the terminal device's support for 5G messages. The detection result may indicate that the terminal fully supports 5G messages, only supports basic SMS, or requires an application client to display rich media content. Based on the detection result, the channel to be used for this push is determined: if the terminal supports 5G messages, the operator's 5G messaging platform is selected as the push channel; if the terminal does not support 5G messages, it automatically switches to a backup channel, such as converting the core message content into structured text and sending it via ordinary SMS, or generating an in-app notification with a link and pushing it to the terminal's notification bar. After determining the push channel, the pre-generated metadata in the interactive 5G message is obtained, and the network adaptation strategy is extracted from it. This strategy contains guidance on how to transmit under specific network conditions, such as prioritizing the transmission of high-priority data blocks when bandwidth is below a threshold. According to the current network conditions and the rules in the strategy, the interactive 5G message is split into multiple data packets, and different transmission priorities are set for data packets of different priorities. Subsequently, the data packets are sent sequentially to the user terminal through the selected push channel.

[0063] During transmission, network conditions and terminal feedback are continuously monitored. If packet loss or delay occurs, retrying or dynamically adjusting the transmission order is performed according to the network adaptation strategy. After receiving the message data packet, the terminal first parses the metadata to obtain the priority markers and compression parameters. The terminal allocates decoding resources according to its own performance parameters and reconstructs the message content hierarchically according to the priority markers: high-priority areas are decoded and presented first, such as displaying the confirmation button and key information text first, ensuring that the user can immediately see the core content and perform operations; then, medium-priority and low-priority areas are decoded and presented step by step in the background, such as loading background images and decorative elements. Finally, the complete interactive 5G message is presented in the call interface of the current call, and the user can intuitively see all the content and start operating.

[0064] It should be noted that by testing the 5G messaging support capabilities of terminal devices and determining the push channel based on the test results, reliable message delivery can be ensured across terminals of different models and configurations, avoiding push failures due to terminal incompatibility. Sending messages based on pre-defined network adaptation strategies within interactive 5G messages allows the transmission process to intelligently adjust according to real-time network conditions, prioritizing the effective delivery of core content. Terminal devices reconstruct the content hierarchically based on priority markers within the message, ensuring that users can see key information and perform actions first, without waiting for all content to load. This process achieves end-to-end collaborative adaptation from sender to receiver, significantly improving message delivery reliability and the immediacy of core information presentation.

[0065] In one embodiment, the step of sending an interactive 5G message to a user's terminal device via a message push channel based on the network adaptation strategy contained in the interactive 5G message includes: splitting the interactive rich media message content of the interactive 5G message into multiple data blocks according to the priority marker contained in the interactive 5G message; and sending each data block to the user's terminal device via a message push channel according to the network adaptation strategy contained in the interactive 5G message.

[0066] A data block is an independent transmission unit formed by splitting interactive rich media message content according to priority markers. Each data block contains one or more information areas with the same priority and is accompanied by a corresponding priority identifier.

[0067] In one implementation, when determining the push channel and preparing to send an interactive 5G message, the message content is first preprocessed before sending. Priority markers are extracted from the metadata of the interactive 5G message. These markers record in detail the hierarchical classification of each information area in the message content. For example, the payment confirmation button and amount display area are marked as high priority, the product image is marked as medium priority, and the background texture is marked as low priority. Based on these priority markers, the entire interactive rich media message content is split into multiple independent data blocks. Each data block contains one or more information areas of the same priority, and a corresponding priority identifier is attached to each data block. For example, a high-priority data block may only contain a confirmation button and amount text, a medium-priority data block contains a product image and description text, and a low-priority data block contains background elements. After splitting, a network adaptation strategy is obtained from the metadata. This strategy contains detailed sending rules, such as sending all data blocks in priority order when the bandwidth is greater than 1Mbps, sending only high-priority and medium-priority data blocks when the bandwidth is between 500Kbps and 1Mbps, sending only high-priority data blocks when the bandwidth is less than 500Kbps, and rules for resuming transmission from the breakpoint after a transmission interruption.

[0068] Real-time monitoring of current network quality parameters, including available bandwidth, round-trip latency, and packet loss rate, is performed. These parameters are then matched against rules in the network adaptation strategy to determine the appropriate transmission scheme under the current network conditions. Based on the matching results, data blocks are sent to the user's terminal device through the previously determined push channel. During transmission, the order and priority determined by the network adaptation strategy are strictly followed. For example, when bandwidth is low, high-priority data blocks are sent first. After the high-priority data blocks are sent and confirmed by the terminal, the decision to continue sending medium-priority data blocks is made based on network conditions. If network interruption or packet loss occurs during transmission, the position of successfully sent data blocks is recorded according to the breakpoint resume rules in the strategy. Once the network recovers, the remaining data blocks are sent from the breakpoint to avoid retransmitting successfully delivered content.

[0069] It's important to note that by dividing interactive rich media message content into multiple data blocks based on priority markers, content of varying importance can be managed and transmitted independently. Dynamically sending each data block according to a network adaptation strategy ensures timely delivery of high-priority data blocks even when network conditions change, guaranteeing users see core information and can perform actions first, while preventing transmission failures due to network fluctuations. This differentiated transmission control effectively improves message delivery success rates in weak network environments, reduces user waiting time, and ensures critical business information is not lost due to network problems.

[0070] In one embodiment, the steps of receiving operation instructions generated by a user interacting with interactive 5G messages on a call interface and synchronizing the operation instructions to an external business system to execute business processing include: capturing the user's interaction behavior with interactive 5G messages on the call interface through an event listening interface, and generating corresponding operation instructions based on the interaction behavior; synchronizing the operation instructions to an external business system through an encrypted channel so that the external business system can execute corresponding business processing according to the operation instructions.

[0071] The event listening interface refers to the programming interface provided by the terminal operating system for capturing user interaction behavior. Through this interface, various user actions on the interactive 5G messages on the call interface can be acquired in real time, such as clicking, swiping, and inputting. Interaction behavior refers to the specific actions taken by the user when interacting with the interactive 5G messages displayed on the call interface, including clicking a confirmation button, filling out an electronic form, selecting options from a list, and swiping to view content. The encrypted channel refers to a data transmission path established using a secure transmission protocol such as TLS, used to ensure the confidentiality and integrity of operation instructions during transmission from the current system to external business systems, preventing data from being eavesdropped on or tampered with.

[0072] In one implementation, after the interactive 5G message is successfully pushed to the call interface and presented to the user, the operation monitoring phase begins. The event monitoring interface provided by the terminal operating system captures various user interactions with the message on the call interface in real time. For example, when the user clicks the "Confirm Payment" button in the message, the event monitoring interface immediately captures the click event and obtains the button's identifier, location information, and associated business data. Based on the captured event type and event source information, a corresponding operation instruction is generated. Taking payment confirmation as an example, the generated instruction includes the operation type as "payment confirmation," the operation object as the order number, the operation data as the specific value of the order number, and the session identifier obtained from the current call context. After generating the instruction, necessary preprocessing is performed, such as verifying the integrity of the instruction, checking whether necessary parameters are complete, and filtering out possible invalid or duplicate operations. After preprocessing, the operation instruction is synchronized in real time to the corresponding external business system through a pre-established encrypted channel. The encrypted channel typically uses the TLS protocol to ensure the security of the instruction during transmission. After receiving the instruction, the external business system first decrypts and parses the instruction to obtain the operation type and operation data, and then executes the corresponding business processing according to the instruction content. For example, in the case of a payment confirmation instruction, the external business system calls the payment interface to complete the deduction operation and records the transaction status. Throughout the entire process, the user can initiate the operation without leaving the call interface, and the call remains connected.

[0073] It's worth noting that by capturing user interactions on the call interface in real time via an event listening interface and generating operation commands, user actions can be instantly transformed into executable business requests without requiring users to manually switch applications or resubmit them afterward. Synchronizing these operation commands to external business systems via an encrypted channel ensures the security and integrity of the commands during transmission, preventing the leakage or tampering of sensitive data. This process allows users to trigger actual business operations on external systems without interrupting the call, achieving a seamless real-time connection from user actions to business processing, while also ensuring the security and reliability of the transaction process.

[0074] The following describes the call service processing apparatus based on 5G messages provided in the embodiments of this application. The call service processing apparatus based on 5G messages described below can be referred to in correspondence with the call service processing method based on 5G messages described above. As shown in Figure 2, this application provides a service processing device for calls based on 5G messages. The device includes: a service scenario determination module 201, used to analyze the collected call content when an authorized user is making a call, determine the service scenario corresponding to the current call, and extract key information corresponding to the service scenario from the call content; an interactive 5G message generation module 202, used to call the interface of the corresponding external business system based on the service scenario and key information, determine the service data associated with the current call, and dynamically generate an interactive 5G message based on the service data, the interactive 5G message containing interactive elements for completing service operations; an interactive 5G message push module 203, used to push the interactive 5G message to the call interface of the current call; and a service processing result update module 204, used to receive the operation instructions generated by the user operating the interactive 5G message on the call interface, synchronize the operation instructions to the external business system to execute service processing, and update and present the processing results returned by the external business system on the call interface.

[0075] In one embodiment, the business scenario determination module 201 includes: a candidate scenario determination unit, used to perform preliminary screening of the collected call content through keyword matching to obtain candidate scenarios; a business scenario determination unit, used to perform semantic analysis on the call content using a deep learning-based semantic understanding model to identify the user's true intent, and determine the business scenario corresponding to the current call from the candidate scenarios based on the identification results; and a key information extraction unit, used to extract key information from the call content using a named entity recognition model combined with entity tags associated with the business scenario.

[0076] In one embodiment, the interactive 5G message generation module 202 includes: an external business system interface determination unit, used to determine the external business system and its API interface corresponding to the business scenario from a pre-configured interface mapping relationship according to the business scenario; a request parameter construction unit, used to construct API request parameters based on key information when the external business system and its API interface pass security verification; and a business data determination unit, used to call the API interface of the external business system according to the API request parameters using a standard protocol, obtain the return data of the external business system, and parse the return data to obtain the business data associated with the current call.

[0077] In one embodiment, the interactive 5G message generation module 202 includes: an interactive message target acquisition unit, used to acquire an interactive message template matching the business scenario from a predefined template library, wherein the interactive message template contains one or more interactive elements; an interactive rich media message content generation unit, used to fill the interactive message template with business data through a dynamic rendering engine to generate interactive rich media message content; a priority division unit, used to analyze the interactive rich media message content and prioritize the identified information regions; a network quality parameter acquisition unit, used to monitor the current network status and obtain network quality parameters; a hierarchical compression processing unit, used to perform hierarchical compression processing on the interactive rich media message content according to the priority of each information region and the network quality parameters, and generate metadata containing compression parameters, priority tags and network adaptation strategies; and an interactive 5G message forming unit, used to encapsulate the hierarchically compressed interactive rich media message content and metadata together to form an interactive 5G message.

[0078] In one embodiment, the interactive 5G message push module 203 includes: a message push channel determination unit, used to detect the 5G message support capability of the user's terminal device and determine the message push channel based on the detection result; and an interactive 5G message push unit, used to send the interactive 5G message to the user's terminal device through the message push channel based on the network adaptation strategy contained in the interactive 5G message, so that the terminal device can perform hierarchical reconstruction of the interactive rich media message content of the interactive 5G message according to the priority marker contained in the interactive 5G message and present it to the call interface of the current call.

[0079] In one embodiment, the interactive 5G message push unit includes: an interactive rich media message content splitting subunit, used to split the interactive rich media message content of the interactive 5G message into multiple data blocks according to the priority marker contained in the interactive 5G message; and a data block sending subunit, used to send each data block to the user's terminal device through the message push channel according to the network adaptation strategy contained in the interactive 5G message.

[0080] In one embodiment, the service processing result update module 204 includes: an operation instruction generation unit, used to capture the user's interaction behavior with interactive 5G messages on the call interface through an event listening interface, and generate corresponding operation instructions based on the interaction behavior; and an operation instruction synchronization unit, used to synchronize the operation instructions to an external business system through an encrypted channel, so that the external business system can perform corresponding service processing according to the operation instructions.

[0081] In one embodiment, this application also provides a storage medium storing computer-readable instructions that, when executed by one or more processors, cause the one or more processors to perform the steps of the 5G message-based call service processing method as described in any of the above embodiments.

[0082] In one embodiment, this application also provides a computer device storing computer-readable instructions that, when executed by one or more processors, cause the one or more processors to perform the steps of the 5G message-based call service processing method as described in any of the above embodiments.

[0083] Schematably, as shown in FIG3, FIG3 is a schematic diagram of the internal structure of a computer device 300 provided in an embodiment of this application. The computer device 300 can be provided as a server. Referring to FIG3, the computer device 300 includes a processing component 302, which further includes one or more processors, and memory resources represented by memory 301 for storing instructions executable by the processing component 302, such as application programs. The application programs stored in memory 301 may include one or more modules, each corresponding to a set of instructions. Furthermore, the processing component 302 is configured to execute instructions to perform the 5G message-based call service processing method of any of the above embodiments.

[0084] The computer device 300 may also include a power supply component 303 configured to perform power management of the computer device 300, a wired or wireless network interface 304 configured to connect the computer device 300 to a network, and an input / output (I / O) interface 305. The computer device 300 may operate on an operating system stored in memory 301, such as Windows Server™, Mac OS X™, Unix™, Linux™, Free BSD™, or similar.

[0085] Those skilled in the art will understand that the structure shown in Figure 3 is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or may combine certain components, or may have different component arrangements.

[0086] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element. In this document, "a," "an," "the," "the," and "its" may also include plural forms unless the context clearly indicates otherwise. "Multiple" refers to at least two, such as 2, 3, 5, or 8, etc. "And / or" includes any and all combinations of the related listed items.

[0087] The various embodiments in this specification are described in a progressive manner. Each embodiment focuses on the differences from other embodiments. The various embodiments can be combined as needed, and the same or similar parts can be referred to each other.

[0088] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for handling services during a call based on 5G messages, characterized in that, The method includes: when an authorized user is in a call, analyzing the collected call content to determine the business scenario corresponding to the current call, and extracting key information corresponding to the business scenario from the call content; based on the business scenario and the key information, calling the interface of the corresponding external business system to determine the business data associated with the current call, and dynamically generating an interactive 5G message based on the business data, the interactive 5G message containing interactive elements for completing business operations; pushing the interactive 5G message to the call interface of the current call; receiving the operation instructions generated by the user operating the interactive 5G message on the call interface, synchronizing the operation instructions to the external business system to execute business processing, and updating and presenting the processing results returned by the external business system on the call interface.

2. The method for handling services during a call based on 5G messages according to claim 1, characterized in that, The steps of analyzing the collected call content, determining the business scenario corresponding to the current call, and extracting key information corresponding to the business scenario from the call content include: performing preliminary screening of the collected call content through keyword matching to obtain candidate scenarios; using a deep learning-based semantic understanding model to perform semantic analysis on the call content, identifying the user's true intent, and determining the business scenario corresponding to the current call from the candidate scenarios based on the identification results; and using a named entity recognition model, combined with entity tags associated with the business scenario, to extract the key information from the call content.

3. The method for handling services during a call based on 5G messages according to claim 1, characterized in that, The step of calling the interface of the corresponding external business system based on the business scenario and the key information to determine the business data associated with the current call includes: determining the external business system and its API interface corresponding to the business scenario from a pre-configured interface mapping relationship; when the external business system and its API interface pass security verification, constructing API request parameters based on the key information; calling the API interface of the external business system using a standard protocol according to the API request parameters, obtaining the return data of the external business system, and parsing the return data to obtain the business data associated with the current call.

4. The method for handling services during a call based on 5G messages according to claim 1, characterized in that, The step of dynamically generating interactive 5G messages based on the business data includes: obtaining an interactive message template matching the business scenario from a predefined template library, wherein the interactive message template contains one or more interactive elements; filling the interactive message template with the business data through a dynamic rendering engine to generate interactive rich media message content; analyzing the interactive rich media message content and prioritizing each identified information region; monitoring the current network status to obtain network quality parameters; performing layered compression processing on the interactive rich media message content according to the priority of each information region and the network quality parameters, and generating metadata containing compression parameters, priority tags, and network adaptation strategies; and encapsulating the layered compressed interactive rich media message content together with the metadata to form the interactive 5G message.

5. The method for handling services during a call based on 5G messages according to claim 1, characterized in that, The step of pushing the interactive 5G message to the call interface of the current call includes: detecting the 5G message support capability of the user's terminal device, determining the message push channel based on the detection result; and sending the interactive 5G message to the user's terminal device through the message push channel based on the network adaptation strategy contained in the interactive 5G message, so that the terminal device can perform hierarchical reconstruction of the interactive rich media message content of the interactive 5G message according to the priority marker contained in the interactive 5G message, and present it to the call interface of the current call.

6. The method for handling services during a call based on 5G messages according to claim 5, characterized in that, The step of sending the interactive 5G message to the user's terminal device through the message push channel based on the network adaptation strategy contained in the interactive 5G message includes: splitting the interactive rich media message content of the interactive 5G message into multiple data blocks according to the priority marker contained in the interactive 5G message; and sending each of the data blocks to the user's terminal device through the message push channel according to the network adaptation strategy contained in the interactive 5G message.

7. The method for handling services during a call based on 5G messages according to claim 1, characterized in that, The step of receiving the operation instructions generated by the user interacting with the interactive 5G message on the call interface and synchronizing the operation instructions to the external business system to perform business processing includes: capturing the user's interaction behavior with the interactive 5G message on the call interface through an event listening interface, and generating corresponding operation instructions based on the interaction behavior; synchronizing the operation instructions to the external business system through an encrypted channel, so that the external business system performs corresponding business processing according to the operation instructions.

8. A call service processing device based on 5G messaging, characterized in that, The device includes: a service scenario determination module, used to analyze the collected call content when an authorized user is making a call, determine the service scenario corresponding to the current call, and extract key information corresponding to the service scenario from the call content; an interactive 5G message generation module, used to call the interface of the corresponding external business system based on the service scenario and the key information, determine the service data associated with the current call, and dynamically generate an interactive 5G message based on the service data, the interactive 5G message containing interactive elements for completing service operations; an interactive 5G message push module, used to push the interactive 5G message to the call interface of the current call; and a service processing result update module, used to receive the operation instructions generated by the user operating the interactive 5G message on the call interface, synchronize the operation instructions to the external business system to execute service processing, and update the processing results returned by the external business system in the call interface.

9. A storage medium, characterized in that: The storage medium stores computer-readable instructions, which, when executed by one or more processors, cause the one or more processors to perform the steps of the 5G message-based call service processing method as described in any one of claims 1 to 7.

10. A computer device, characterized in that, include: One or more processors, and a memory; the memory stores computer-readable instructions that, when executed by the one or more processors, perform the steps of the 5G message-based call service processing method as described in any one of claims 1 to 7.

Images