AI modular digital communication method, device and equipment based on low-code platform

Abstract

The invention relates to an AI modular digital communication method, device and equipment based on a low-code platform, and belongs to the technical field of low-code platforms, and the method comprises the steps: collecting the operation data of each independent service plate of the low-code platform, recognizing the high-frequency composite operation information in a dynamic behavior feature map, generating a cross-service plate association rule set, and carrying out the recognition of the high-frequency composite operation information in a dynamic behavior feature map; creating a communication link across service plates through the association rule set; when a user demand instruction is obtained, triggering a code generation engine of an AI module, and creating a metadata architecture of a new function module under the corresponding service plate; a two-channel verification mechanism is adopted to carry out sandbox testing on the new function module, communication link updating is carried out on a service plate with the new function module, and updating comprises adjustment of an association rule set; therefore, the operation and maintenance complexity and upgrading cost of the system are reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of low-code platforms, and in particular to an AI modular digital communication method, device and equipment based on a low-code platform. Background Art

[0002] While current low-code platforms enable rapid deployment of business modules, traditional approaches suffer from significant drawbacks: inter-module communication links rely on manual configuration, making it difficult to dynamically adapt to changing business rules; new feature development still requires coding, lacking intelligent iteration capabilities based on user behavior; and system stability verification is limited to routine scenario testing, unable to effectively handle sudden traffic bursts or complex fault chains. Existing low-code platforms have yet to emerge that can deeply integrate AI self-learning mechanisms, dynamically optimize cross-module communication links, and autonomously defend against extreme scenarios. This leaves enterprises facing issues such as system rigidity, high operational and maintenance costs, and insufficient fault tolerance. Summary of the Invention

[0003] To achieve the above objectives, the present invention provides an AI modular digital communication method based on a low-code platform, comprising the following steps:

[0004] Collecting operational data from each independent business segment of the low-code platform, including interface operation traces, API call frequency, and data processing exception logs, and inputting the operational data into the AI model training platform to construct a dynamic behavior feature map;

[0005] Identify high-frequency complex operation information in the dynamic behavior feature map, generate a cross-business segment association rule set, and create a cross-business segment communication link through the association rule set;

[0006] When a user demand instruction is obtained, the code generation engine of the AI module is triggered to create a metadata architecture for a new functional module under the corresponding business segment;

[0007] A dual-channel verification mechanism is used to conduct sandbox testing on the new functional modules. The first channel verifies functional integrity by replaying historical business data, and the second channel simulates platform stability in extreme scenarios based on Monte Carlo tree search.

[0008] The communication link of the business segment having the new functional module is updated, and the updating includes adjusting the association rule set.

[0009] Furthermore, the step of identifying high-frequency composite operation information in the dynamic behavior characteristic map includes:

[0010] Perform multi-dimensional slice analysis on the dynamic behavior feature map and divide the sequence segments of the interface operation trajectory based on the time window;

[0011] A density clustering algorithm is used to identify high-density clustered areas of the sequence segments, and a hidden Markov model is combined to extract temporal dependencies across business segments to obtain candidate patterns with a composite operation set.

[0012] The support and confidence of candidate patterns are calculated using an association rule mining algorithm. Patterns that meet the threshold are marked as high-frequency composite operation information. The confidence calculation introduces business weight factors, including API call frequency weight and abnormal log relevance weight.

[0013] The marked high-frequency composite operation information is subjected to pattern verification, wherein the pattern verification includes matching a historical high-frequency operation pattern library through cosine similarity, and screening out candidate patterns with confidence differences exceeding a preset value as effectively discovered new composite operation information.

[0014] Furthermore, the step of generating a cross-business segment association rule set and creating a cross-business segment communication link using the association rule set includes:

[0015] Generate a cross-business segment association rule set based on the high-frequency composite operation information, wherein the association rule set includes data interaction logic and trigger conditions between business modules;

[0016] Automatically create cross-business segment communication links based on the association rule set, including configuring data transmission protocols, interface connection parameters, and exception handling strategies;

[0017] The effectiveness and stability of the communication link are verified through an automated testing environment, and the link configuration parameters are dynamically optimized based on the verification results.

[0018] Furthermore, when a user demand instruction is obtained, the code generation engine of the AI module is triggered to create a metadata architecture of a new functional module under the corresponding business segment, including:

[0019] Perform semantic analysis on user demand instructions to extract functional target parameters and associated business segment identifiers;

[0020] Calling an AI code generation engine based on the functional target parameters to generate a metadata architecture template including data structure definition, interface constraints, and permission configuration;

[0021] Verify the compatibility of the metadata architecture with existing business segments through a simulated test environment, and dynamically adjust interface constraints based on the verification results;

[0022] Inject the optimized metadata architecture into the functional module library of the target business segment and update the dependency graph between modules.

[0023] Furthermore, the first channel verifies functional integrity by replaying historical service data, including:

[0024] Extract time-stamped business data samples from the historical database of the target business segment and construct a test data set based on preset screening conditions;

[0025] Injecting the test data set into the execution environment of the new functional module to trigger the complete life cycle operation process of the module interface;

[0026] Comparing module output data with expected results of corresponding tags in historical business data to generate a functional integrity verification report, which includes quantitative indicators of interface response deviation and logic execution coverage;

[0027] The module functional integrity status is determined based on the deviation threshold in the verification report. If it does not meet the standard, it will automatically roll back to the previous version of the module architecture and trigger the AI model to regenerate the metadata template.

[0028] Furthermore, the second channel simulates the platform stability under extreme scenarios based on Monte Carlo tree search, including:

[0029] Generate an extreme scenario parameter set based on the operational characteristics of the business segment, including the concurrent request peak, network latency fluctuation range, and service degradation trigger conditions;

[0030] The Monte Carlo tree search algorithm is used to simulate the multi-level fault propagation path under the extreme scenario parameters and calculate the quantitative stability indicators of the platform's core services.

[0031] When the quantitative indicator is lower than a preset threshold, the platform configuration is automatically adjusted.

[0032] Furthermore, the step of updating the communication link of the business segment having the new functional module includes:

[0033] Detecting interface feature data of new functional modules, including input and output data structure, protocol compatibility range, and historical call frequency;

[0034] Perform matching analysis on the interface feature data and the communication rules of the existing business segments, and update the data interaction logic in the association rule set based on a preset matching strategy;

[0035] Generate updated communication link configuration parameters, including adapted data transmission protocols, interface mapping relationships, and exception handling mechanisms;

[0036] Deploy the updated communication link configuration through the automated testing platform to verify the connectivity between the new functional module and the associated business segments;

[0037] The communication link configuration parameters are dynamically adjusted according to the verification results, and the dependency graph between platform modules is synchronously updated.

[0038] The present invention proposes an AI modular digital communication device based on a low-code platform, comprising:

[0039] A collection unit is used to collect operating data of each independent business segment of the low-code platform, the operating data including interface operation traces, API call frequency and data processing exception logs, and input the operating data into the AI model training platform to construct a dynamic behavior feature map;

[0040] A communication unit, configured to identify high-frequency composite operation information in a dynamic behavior feature map, generate a cross-business segment association rule set, and establish a cross-business segment communication link through the association rule set;

[0041] The AI unit is used to trigger the code generation engine of the AI module when obtaining user demand instructions, and create a metadata architecture of the new functional module under the corresponding business segment;

[0042] A testing unit, configured to perform sandbox testing on the new functional module using a dual-channel verification mechanism, wherein the first channel verifies functional integrity by replaying historical business data, and the second channel simulates platform stability under extreme scenarios based on Monte Carlo tree search;

[0043] An updating unit is configured to update a communication link of a business segment having the new functional module, wherein the updating includes adjusting an association rule set.

[0044] The present invention also provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, it implements the steps of the above-mentioned AI modular digital communication method based on a low-code platform.

[0045] The present invention also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of the above-mentioned AI modular digital communication method based on a low-code platform are implemented.

[0046] The AI modular digital communication method, device, and equipment based on a low-code platform provided by the present invention have the following beneficial effects:

[0047] Automatically generate communication links across business modules based on user behavior characteristics to reduce manual configuration workload; improve platform stability and shorten fault recovery time through a dual-channel verification system (historical data playback + extreme scenario simulation); dynamic module combination recommendation mechanism speeds up platform function iteration, meeting the personalized needs of enterprises while maintaining the unity of the core architecture, significantly reducing system operation and maintenance complexity and upgrade costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 This is a flowchart of an AI modular digital communication method based on a low-code platform in one embodiment of the present invention;

[0049] Figure 2 1 is a flow chart of an AI modular digital communication method based on a low-code platform in another embodiment of the present invention;

[0050] Figure 3 It is a flowchart of an AI modular digital communication method based on a low-code platform in another embodiment of the present invention;

[0051] Figure 4 Schematic diagram of the principle of an AI modular digital communication method based on a low-code platform in one embodiment of the present invention;

[0052] Figure 5 This is a structural block diagram of an AI modular digital communication device based on a low-code platform in one embodiment of the present invention;

[0053] Figure 6 It is a schematic block diagram of the structure of a computer device according to an embodiment of the present invention.

[0054] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0055] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0056] Reference Figure 1 , which is a flowchart of an AI modular digital communication method based on a low-code platform. The method includes:

[0057] S1: Collect the operating data of each independent business segment of the low-code platform, including interface operation traces, API call frequency, and data processing exception logs, and input the operating data into the AI model training platform to construct a dynamic behavior feature map;

[0058] S2, identifying high-frequency composite operation information in the dynamic behavior feature map, generating a cross-business segment association rule set, and establishing a cross-business segment communication link through the association rule set;

[0059] S3, when receiving user demand instructions, triggers the code generation engine of the AI module to create the metadata architecture of the new functional module under the corresponding business segment;

[0060] S4: Sandbox testing of the new functional module using a dual-channel verification mechanism. The first channel verifies functional integrity by replaying historical business data, and the second channel simulates platform stability under extreme scenarios based on Monte Carlo tree search.

[0061] S5, updating the communication link of the business segment having the new functional module, wherein the updating includes adjusting the association rule set.

[0062] In one embodiment, in step S1 above, a visual tracking probe is deployed on the platform front-end interface, using DOM event monitoring technology to capture the spatiotemporal feature data of user operation trajectories. A lightweight monitoring middleware is also embedded in the back-end service layer to collect API call frequency and exception logs in a non-invasive manner. The data collection process is shown in Table 1:

[0063]

[0064]

[0065] Table 1

[0066] The above dynamic behavior feature map is generated using the AI model training platform. The generation method is as follows:

[0067] First, generate vectors corresponding to the interface operation trajectory, API call frequency, and exception log, respectively, and embed the operation sequence vector V. op ∈R 128 , frequency domain eigenvector V api ∈R 64 , semantic vector V err ∈R 32 , and establish a joint coding formula V based on the three types of vectors node =ReLU(W·[V op ||V api ||]V err +b), where W is the trainable weight matrix and b is the loss function algorithm.

[0068] The dynamic graph convolutional network DGCN is used to perform convolution processing on the combined three types of vectors, and the node representation is iteratively updated through three layers of DGCN, and finally the dynamic behavior feature map of the module association degree is output.

[0069] Reference Attachment Figure 2 In one embodiment, in step S2 of identifying high-frequency composite operation information in a dynamic behavior feature map, generating a cross-business segment association rule set, and creating a cross-business segment communication link using the association rule set, step S21 of identifying high-frequency composite operation information in a dynamic behavior feature map includes:

[0070] S211. Perform multi-dimensional slicing analysis on the dynamic behavior feature map, and divide the sequence segments of the interface operation trajectory based on the time window;

[0071] Specifically, the multi-dimensional slicing analysis of the dynamic behavior feature map disassembles the user operation behavior through time and business dimensions. The time window division adopts an adaptive method, dynamically adjusting the window boundary according to the operation interval duration and semantic relevance to ensure the business logic coherence of the operation sequence within each time segment. For example, the consecutive contract approval operations during the end-of-month settlement period will be automatically clustered into the same analysis unit.

[0072] S212. Use the density clustering algorithm to identify the high-density aggregation regions of the sequence segments, and combine the hidden Markov model to extract the temporal dependence relationship across business segments to obtain candidate patterns with a composite operation set;

[0073] Specifically, use density clustering (DBSCAN) to identify high-density operation sequences. The identification method is as follows: Set the sequence segment radius C and the minimum number of samples min, and calculate for each data point x of the sample i within the domain of radius C, through the formula N c (x i ) = {x j ∈ X | dist(x i , x j ) < C}. If it is judged that N c (x i ) < min, then the data point x i is marked as the high-density region of the sequence segment, and finally a high-density data point set is formed; then, extract the cross-business temporal dependence relationship through the hidden Markov model (HMM), and observe the relevance of the high-density regions in the sequence segment through HMM. For example, if HMM finds that "contract approval completed" is usually followed by "financial posting", then the operations of these two business modules form a candidate composite pattern.

[0074] S213. Calculate the support and confidence of the candidate patterns through the association rule mining algorithm, and mark the patterns that meet the threshold as high-frequency composite operation information, where the confidence calculation introduces business weight factors, including the API call frequency weight and the abnormal log correlation weight;

[0075] Specifically, the association rule mining algorithm uses the Apriori algorithm for improvement to calculate the support and confidence of the candidate patterns on the platform. The support calculation is as follows:

[0076]

[0077] Among them, X → Y represents the set of user operations in the candidate pattern, and N is the total number of transactions.

[0078] The confidence is calculated as:

[0079]

[0080] Among them, w api is the API call frequency weight, w Error is the abnormal log correlation weight.

[0081] S214 , performing pattern verification on the marked high-frequency composite operation information, wherein the pattern verification includes matching a historical high-frequency operation pattern library through cosine similarity, and screening out candidate patterns with confidence differences exceeding a preset value as effectively discovered new composite operation information.

[0082] Specifically, after identifying candidate high-frequency composite operation patterns through association rule mining, further verification is required to determine whether they are truly effective new business discoveries (rather than variations of existing historical patterns). This verification method utilizes cosine similarity matching combined with confidence difference analysis to ensure the uniqueness and business value of the new pattern.

[0083] The candidate pattern P to be verified new The eigenvector V new Perform dashed line similarity matching calculation using the algorithm:

[0084]

[0085] θ sim The historical patterns exceeding a threshold (preferably 85%) are recorded as a similar historical pattern set Psim.

[0086] The confidence of the candidate pattern is compared with that of similar historical patterns. If the difference between the comparisons is greater than a threshold (preferably 15%), it is determined to be newly discovered composite operation information.

[0087] Reference Attachment Figure 3 In one embodiment, in step S2 of identifying high-frequency composite operation information in a dynamic behavior feature graph, generating a cross-business segment association rule set, and creating a cross-business segment communication link using the association rule set, step S22 of generating a cross-business segment association rule set and creating a cross-business segment communication link using the association rule set includes:

[0088] S221, generating a cross-business segment association rule set based on the high-frequency composite operation information, wherein the association rule set includes data interaction logic and trigger conditions between business modules;

[0089] Convert high-frequency complex operation information (such as "contract approval → financial accounting") into If-Then rules, introduce business weight factors (such as the financial interface call priority is P1), dynamically adjust the rule triggering order, and generate a rule set for data association between low-code platforms.

[0090] S222, automatically creating a cross-business segment communication link based on the association rule set, including configuring a data transmission protocol, interface connection parameters, and an exception handling strategy;

[0091] The low-code platform directly establishes protocol adaptation for each business segment and automatically selects the transmission protocol based on the interface document of the target business module (such as Swagger), specifically:

[0092] Interface type default protocol selection logic High real-time requirements gRPC evaluation delay < 100ms enabled

[0093] Cross-platform interaction REST / JSON is selected based on the caller SDK compatibility

[0094] S223, verifying the effectiveness and stability of the communication link through an automated test environment, and dynamically optimizing link configuration parameters based on the verification results.

[0095] The end-to-end link is verified using the forward and reverse tests of the closed-loop verification system. The forward test includes checking whether the financial module generates corresponding accounting records, and the reverse test includes simulating interface timeouts to verify whether the retry mechanism is triggered, thereby determining validity.

[0096] Stability is tested through stability stress testing: using the Locust tool to simulate a stepped load (from 100 to 1000 TPS) and monitoring indicators:

[0097] Indicator type threshold optimization strategy

[0098] Average response time > 200ms Switch protocol (HTTP / 1.1 → gRPC)

[0099] If the error rate is > 5%, scale up the container instance.

[0100] In one embodiment, when a user demand instruction is obtained, the code generation engine of the AI module is triggered to create a metadata architecture of a new functional module under the corresponding business segment, including the following steps:

[0101] Perform semantic analysis on user demand instructions to extract functional target parameters and associated business segment identifiers;

[0102] Calling an AI code generation engine based on the functional target parameters to generate a metadata architecture template including data structure definition, interface constraints, and permission configuration;

[0103] Verify the compatibility of the metadata architecture with existing business segments through a simulated test environment, and dynamically adjust interface constraints based on the verification results;

[0104] Inject the optimized metadata architecture into the functional module library of the target business segment and update the dependency graph between modules.

[0105] During the specific implementation process, in the development of low-code platforms, users face two major technical bottlenecks when adding new functional modules: 1. Traditional demand analysis relies on manual disassembly, which has semantic understanding deviations; 2. Verification of the compatibility of module architecture with existing systems is time-consuming; this embodiment uses AI semantic understanding and intelligent verification mechanisms to achieve automated conversion of requirements to code architecture.

[0106] Receive natural language demand instructions (text / speech-to-text), for example: "Add automatic contract amount verification function to the financial approval section, which requires connection to the ERP system." Use the improved Bi-LSTM-CRF model for entity recognition. The sample code is as follows:

[0107]

[0108] The metadata template generation uses the GPT-3.5-turbo fine-tuning model to generate the initial architecture, as shown below:

[0109]

[0110] In a real example:

[0111] Data Structure

[0112]

[0113] #Interface constraints

[0114]

[0115] #Permission configuration

[0116]

[0117] In the step of verifying the compatibility of the metadata architecture with the existing business segments through a simulated test environment and dynamically adjusting the interface constraints based on the verification results,

[0118] Build a simulation test environment to test the compatibility of metadata architecture with existing business segments. Figure 4, inject simulated data streams into the new module to test data compatibility, interface stability, and permission penetration. The relationship between these and the test indicators is shown in the following table:

[0119]

[0120] The dynamic adjustment mechanism is to automatically trigger constraint optimization when it is detected that the interface response timeout is greater than 500ms.

[0121] #Interface constraint adjustment example:

[0122]

[0123] The steps to inject the optimized metadata architecture into the functional module library of the target business segment and update the dependency graph between modules include:

[0124] Use Neo4j graph database to maintain module relationships.

[0125] In one embodiment, the step of verifying the functional integrity of the first channel by replaying historical service data includes:

[0126] Extract time-stamped business data samples from the historical database of the target business segment and construct a test data set based on preset screening conditions;

[0127] Injecting the test data set into the execution environment of the new functional module to trigger the complete life cycle operation process of the module interface;

[0128] Comparing module output data with expected results of corresponding tags in historical business data to generate a functional integrity verification report, which includes quantitative indicators of interface response deviation and logic execution coverage;

[0129] The module functional integrity status is determined based on the deviation threshold in the verification report. If it does not meet the standard, it will automatically roll back to the previous version of the module architecture and trigger the AI model to regenerate the metadata template.

[0130] In the process of specific implementation,

[0131] The above steps of historical data sample extraction and test set construction are:

[0132] Extract real business data with time stamps from the historical database of the target business segment to build a test data set with high coverage. In the specific technical implementation, first, through the time window screening mechanism, select historical data covering the complete business cycle (such as transaction records in the past 30 days) to ensure coverage of business peaks, troughs and abnormal scenarios. The key fields of the data sample include operation timestamps, input parameters, execution results and error logs. The tags can retain the temporal relationship and contextual dependencies of the original business events. Through preset screening conditions (such as data distribution balance and scenario diversity), a clustering algorithm is used to automatically identify typical business patterns, eliminate redundant data, and finally generate a representative data set that can reflect the actual historical situation and has test efficiency. The key innovation of this process lies in the dynamic alignment technology of timestamps to ensure that the time logic of the test data is strictly consistent with the production environment.

[0133] The above test data injection and life cycle triggering steps are:

[0134] In an isolated sandbox environment, the test data set is injected into the new functional module through a high-precision data playback engine. The engine uses a distributed message queue (such as Kafka) to achieve real-time transmission of data streams, strictly reproducing the order of business events according to the original time interval, with the error controlled at the millisecond level. During the injection process, the complete life cycle of the module is triggered, including initialization configuration, interface call, data processing and result output stages. The environment simulation layer dynamically builds virtual instances of dependent services (such as simulated ERP) to ensure that the module can still execute end-to-end business processes without real external dependencies. The technical core of this stage lies in the spatiotemporal consistency maintenance algorithm, which processes out-of-order data through the event time watermark mechanism to ensure that the time semantics of the test process are completely consistent with the historical scenario.

[0135] The steps for comparing and quantifying the above results are as follows:

[0136] A multi-dimensional comparison engine analyzes the differences between new module output results and historical expectations. At the data structure level, a pattern matching algorithm verifies the compatibility of JSON Schema and database table structures. At the numerical level, statistical deviations of key fields (such as the amount error rate and latency standard deviation) are calculated. At the business logic level, the execution path coverage of decision branches is tracked. The deviation model integrates indicators such as numerical differences, enumerated field matching, and structural integrity, and uses a weighted calculation to derive an overall deviation coefficient. Simultaneously, code instrumentation technology monitors the module's logical execution flow in real time, generating a branch coverage heat map and identifying untriggered exception handling paths. The resulting verification report not only includes quantitative metrics (such as "interface response deviation ≤ 3%") but also identifies specific deviation points (such as "inconsistent amount rounding rules"), providing direct evidence for problem location.

[0137] The above steps of intelligent decision-making and automatic optimization are:

[0138] Functional integrity is automatically determined based on preset quality thresholds (such as overall deviation ≤ 5%, logical coverage ≥ 95%). If the verification fails to meet the standards, a cascade response mechanism is triggered: first, the module is rolled back to the previous stable version through version control (such as Git) to ensure business continuity; at the same time, a knowledge base of failure cases is built to extract key features (such as boundary condition errors caused by specific parameter combinations) and feed them back to the AI training platform. The AI model is based on an incremental learning strategy. While retaining the original knowledge, it focuses on optimizing the metadata generation logic related to newly discovered defects. This process introduces an adversarial training mechanism to automatically generate targeted test cases, giving priority to verifying historical weak links in the next round of iterations to form a self-reinforcing verification loop.

[0139] In one embodiment, the second channel simulates platform stability under extreme scenarios based on Monte Carlo tree search, including:

[0140] Generate an extreme scenario parameter set based on the operational characteristics of the business segment, including the concurrent request peak, network latency fluctuation range, and service degradation trigger conditions;

[0141] The Monte Carlo tree search algorithm is used to simulate the multi-level fault propagation path under the extreme scenario parameters and calculate the quantitative stability indicators of the platform's core services.

[0142] When the quantitative indicator is lower than a preset threshold, the platform configuration is automatically adjusted.

[0143] During the specific implementation process, the stability verification of extreme scenarios based on Monte Carlo tree search first analyzes historical operation data and system bottleneck characteristics to dynamically build an extreme scenario parameter library covering multi-dimensional damage factors:

[0144]

[0145] Then, the system state space is modeled using a tree structure to intelligently explore high-risk fault propagation paths. The tree node definitions in the tree structure include: service health vectors (such as API availability and number of DB connections), environmental parameters (current concurrency and network latency), and path history (triggered fault chains).

[0146] In one embodiment, the step of updating the communication link of the business segment having the new functional module includes:

[0147] Detecting interface feature data of new functional modules, including input and output data structure, protocol compatibility range, and historical call frequency;

[0148] Perform matching analysis on the interface feature data and the communication rules of the existing business segments, and update the data interaction logic in the association rule set based on a preset matching strategy;

[0149] Generate updated communication link configuration parameters, including adapted data transmission protocols, interface mapping relationships, and exception handling mechanisms;

[0150] Deploy the updated communication link configuration through the automated testing platform to verify the connectivity between the new functional module and the associated business segments;

[0151] The communication link configuration parameters are dynamically adjusted according to the verification results, and the dependency graph between platform modules is synchronously updated.

[0152] During implementation, the API definition file for the new module (e.g., the OpenAPI specification) is parsed to extract metadata such as the request / response field type, format constraints (e.g., numeric range, regular expression), and nested structure. The module's supported communication protocols (e.g., HTTP / 2, gRPC, WebSocket) and version numbers are identified, and the ability to convert message encapsulation formats (JSON / Protobuf) is analyzed. The module's interface call frequency distribution during the testing phase is statistically analyzed to predict the peak load pressure in the production environment (e.g., QPS = 200 ± 50).

[0153] Reference Attachment Figure 5 This is a structural block diagram of an AI modular digital communication device based on a low-code platform proposed in the present invention, which includes:

[0154] A collection unit is used to collect operating data of each independent business segment of the low-code platform, the operating data including interface operation traces, API call frequency and data processing exception logs, and input the operating data into the AI model training platform to construct a dynamic behavior feature map;

[0155] A communication unit, configured to identify high-frequency composite operation information in a dynamic behavior feature map, generate a cross-business segment association rule set, and establish a cross-business segment communication link through the association rule set;

[0156] The AI unit is used to trigger the code generation engine of the AI module when obtaining user demand instructions, and create a metadata architecture of the new functional module under the corresponding business segment;

[0157] A testing unit, configured to perform sandbox testing on the new functional module using a dual-channel verification mechanism, wherein the first channel verifies functional integrity by replaying historical business data, and the second channel simulates platform stability under extreme scenarios based on Monte Carlo tree search;

[0158] An updating unit is configured to update a communication link of a business segment having the new functional module, wherein the updating includes adjusting an association rule set.

[0159] Reference Figure 6 In an embodiment of the present invention, a computer device is also provided. The computer device may be a server, and its internal structure may be as follows: Figure 6 As shown. The computer device includes a processor, memory, display screen, input device, network interface and database connected via a system bus. The processor of the computer design is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The database of the computer device is used to store the corresponding data in this embodiment. The network interface of the computer device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, the above method is implemented.

[0160] Those skilled in the art will understand that Figure 6 The structure shown in the figure is merely a block diagram of a portion of the structure related to the solution of the present invention and does not constitute a limitation on the computer device to which the solution of the present invention is applied.

[0161] An embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon, which implements the above-described method when executed by a processor. It is understood that the computer-readable storage medium in this embodiment can be a volatile readable storage medium or a non-volatile readable storage medium.

[0162] Those skilled in the art will appreciate that all or part of the processes in the above-described method embodiments can be implemented by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the above-described method embodiments. Among them, any reference to memory, storage, database, or other media provided by the present invention and used in the embodiments can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double-speed SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct RAM bus dynamic RAM (DRDRAM), and RAMbus dynamic RAM.

[0163] It should be noted that, in this document, the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, apparatus, article, or method comprising a series of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, apparatus, article, or method. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, apparatus, article, or method comprising the element.

[0164] The above description is only a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made by using the contents of the present invention description and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. An AI modular digital communication method based on a low-code platform, characterized in that: The following steps are involved: Collecting operational data from each independent business segment of the low-code platform, including interface operation traces, API call frequency, and data processing exception logs, and inputting the operational data into the AI model training platform to construct a dynamic behavior feature map; Identify high-frequency complex operation information in the dynamic behavior feature map, generate a cross-business segment association rule set, and create a cross-business segment communication link through the association rule set; When a user demand instruction is obtained, the code generation engine of the AI module is triggered to create a metadata architecture for a new functional module under the corresponding business segment; A dual-channel verification mechanism is used to conduct sandbox testing on the new functional modules. The first channel verifies functional integrity by replaying historical business data, and the second channel simulates platform stability in extreme scenarios based on Monte Carlo tree search. The communication link of the business segment having the new functional module is updated, and the updating includes adjusting the association rule set.

2. The AI modular digital communication method based on a low-code platform according to claim 1 is characterized in that: The steps of identifying high-frequency composite operation information in the dynamic behavior characteristic map include: Perform multi-dimensional slice analysis on the dynamic behavior feature map and divide the sequence segments of the interface operation trajectory based on the time window; A density clustering algorithm is used to identify high-density clustered areas of the sequence segments, and a hidden Markov model is combined to extract temporal dependencies across business segments to obtain candidate patterns with a composite operation set. The support and confidence of candidate patterns are calculated using an association rule mining algorithm. Patterns that meet the threshold are marked as high-frequency composite operation information. The confidence calculation introduces business weight factors, including API call frequency weight and abnormal log relevance weight. The marked high-frequency composite operation information is subjected to pattern verification, wherein the pattern verification includes matching a historical high-frequency operation pattern library through cosine similarity, and screening out candidate patterns with confidence differences exceeding a preset value as effectively discovered new composite operation information.

3. The AI modular digital communication method based on a low-code platform according to claim 1 is characterized in that: The steps of generating a cross-business segment association rule set and creating a cross-business segment communication link using the association rule set include: Generate a cross-business segment association rule set based on the high-frequency composite operation information, wherein the association rule set includes data interaction logic and trigger conditions between business modules; Automatically create cross-business segment communication links based on the association rule set, including configuring data transmission protocols, interface connection parameters, and exception handling strategies; The effectiveness and stability of the communication link are verified through an automated testing environment, and the link configuration parameters are dynamically optimized based on the verification results.

4. The AI modular digital communication method based on a low-code platform according to claim 1 is characterized in that: When a user demand instruction is obtained, the code generation engine of the AI module is triggered to create a metadata architecture for a new functional module under the corresponding business segment, including the following steps: Perform semantic analysis on user demand instructions to extract functional target parameters and associated business segment identifiers; Calling an AI code generation engine based on the functional target parameters to generate a metadata architecture template including data structure definition, interface constraints, and permission configuration; Verify the compatibility of the metadata architecture with existing business segments through a simulated test environment, and dynamically adjust interface constraints based on the verification results; Inject the optimized metadata architecture into the functional module library of the target business segment and update the dependency graph between modules.

5. The AI modular digital communication method based on a low-code platform according to claim 1 is characterized in that: The first channel verifies functional integrity by replaying historical business data, including: Extract time-stamped business data samples from the historical database of the target business segment and construct a test data set based on preset screening conditions; Injecting the test data set into the execution environment of the new functional module to trigger the complete life cycle operation process of the module interface; Comparing module output data with expected results of corresponding tags in historical business data to generate a functional integrity verification report, which includes quantitative indicators of interface response deviation and logic execution coverage; The module functional integrity status is determined based on the deviation threshold in the verification report. If it does not meet the standard, it will automatically roll back to the previous version of the module architecture and trigger the AI model to regenerate the metadata template.

6. The AI modular digital communication method based on a low-code platform according to claim 1 is characterized in that: The second channel simulates platform stability in extreme scenarios based on Monte Carlo tree search, including the following steps: Generate an extreme scenario parameter set based on the operational characteristics of the business segment, including the concurrent request peak, network latency fluctuation range, and service degradation trigger conditions; The Monte Carlo tree search algorithm is used to simulate the multi-level fault propagation path under the extreme scenario parameters and calculate the quantitative stability indicators of the platform's core services. When the quantitative indicator is lower than a preset threshold, the platform configuration is automatically adjusted.

7. The AI modular digital communication method based on a low-code platform according to claim 1 is characterized in that: The step of updating the communication link of the business segment having the new functional module includes: Detecting interface feature data of new functional modules, including input and output data structure, protocol compatibility range, and historical call frequency; Perform matching analysis on the interface feature data and the communication rules of the existing business segments, and update the data interaction logic in the association rule set based on a preset matching strategy; Generate updated communication link configuration parameters, including adapted data transmission protocols, interface mapping relationships, and exception handling mechanisms; Deploy the updated communication link configuration through the automated testing platform to verify the connectivity between the new functional module and the associated business segments; The communication link configuration parameters are dynamically adjusted according to the verification results, and the dependency graph between platform modules is synchronously updated.

8. An AI modular digital communication device based on a low-code platform, characterized in that: include: A collection unit is used to collect operating data of each independent business segment of the low-code platform, the operating data including interface operation traces, API call frequency and data processing exception logs, and input the operating data into the AI model training platform to construct a dynamic behavior feature map; A communication unit, configured to identify high-frequency composite operation information in a dynamic behavior feature map, generate a cross-business segment association rule set, and establish a cross-business segment communication link through the association rule set; The AI unit is used to trigger the code generation engine of the AI module when obtaining user demand instructions, and create a metadata architecture of the new functional module under the corresponding business segment; A testing unit, configured to perform sandbox testing on the new functional module using a dual-channel verification mechanism, wherein the first channel verifies functional integrity by replaying historical business data, and the second channel simulates platform stability under extreme scenarios based on Monte Carlo tree search; An updating unit is configured to update a communication link of a business segment having the new functional module, wherein the updating includes adjusting an association rule set.

9. A computer device comprising a memory and a processor, wherein a computer program is stored in the memory, wherein: When the processor executes the computer program, it implements the steps of the AI modular digital communication method based on the low-code platform as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the AI modular digital communication method based on a low-code platform described in any one of claims 1 to 7 are implemented.

Images