Real-time data updating method and device based on AI prediction and computer equipment

Abstract

The invention provides a real-time data updating method and device based on AI prediction and computer equipment. The method comprises the steps that a server side collects a historical behavior data set of a user to train an LSTM time sequence model; the client establishes a long connection with the server, and reports the current operation behavior of the user to the server in real time; the server predicts the next operation behavior of the user based on the LSTM time sequence model; the server side compares a local version number carried by the client side with a server version number by maintaining the version number of the data item, adopts a Diff algorithm to calculate incremental data, and issues full data if the incremental data volume exceeds a full data threshold value; compressing the incremental data by using a compression algorithm, and actively issuing the incremental data to a client through long connection; and the client side merges the incremental data and then executes MD5 verification, and if the verification is inconsistent, the client side requests to update the total data. The method has the beneficial effect that the data updating efficiency and reliability in the long connection environment are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data update, and in particular to a real-time data update method, device and computer equipment based on AI prediction. Background Art

[0002] With the development of information technology, the demand for real-time data update in various application programs is increasing day by day. In the traditional data update method, data is not prepared in advance according to the usage characteristics of different user APPs, and full-volume data transmission is often adopted between the client and the server, which leads to problems such as bandwidth waste, increased latency and resource occupation.

[0003] The APP page or function data is basically updated with the server after the user enters the page or executes a certain operation. Due to network or the amount of updated data, it may cause slow data requests, resulting in the user not being able to see the updated data page in time, affecting the user experience. Summary of the Invention

[0004] The technical problem to be solved by the present invention is: to provide a real-time data update method, device and computer equipment based on AI prediction, aiming to improve the data update efficiency and reliability in the long connection environment.

[0005] To solve the above technical problem, the technical solution adopted by the present invention is: a real-time data update method based on AI prediction, including the following steps:

[0006] The server collects the historical behavior data set of the user, and the historical behavior data includes the user's access frequency, time period preference, browsing data, geographical features, age information, the APP page currently accessed by the user, the button clicked on the user's current page, the user's page operation path and events;

[0007] Use the LSTM time series model to train the historical behavior data set, divide the historical behavior data set into a training set and a test set according to a ratio, and optimize the LSTM time series model by adjusting the hidden layer dimension, the number of LSTM layers or the learning rate parameter until the prediction accuracy of the test set exceeds 70%;

[0008] The client establishes a long connection with the server and reports the user's current operation behavior to the server in real time;

[0009] The server predicts the user's next operation behavior based on the LSTM time series model and prepares the page or function data corresponding to the predicted operation in advance;

[0010] The server maintains the version numbers of data items, compares the local version number carried by the client with the server version number, calculates the incremental data using the Diff algorithm, and if the amount of incremental data exceeds the full data threshold, it sends the full data;

[0011] Use a compression algorithm to compress the incremental data and actively send it to the client through a long connection;

[0012] After the client merges the incremental data, it performs an MD5 check. If the check is inconsistent, it requests a full data update.

[0013] Furthermore, the training of the LSTM time series model also includes:

[0014] Regularly add the newly generated user behavior data to the training set to dynamically optimize the LSTM time series model;

[0015] Train through the LSTM time series model built into the PyTorch tool, and only adjust a single parameter each time to optimize the model performance.

[0016] Furthermore, the step of using a compression algorithm to compress the incremental data and actively send it to the client through a long connection specifically includes:

[0017] If the client requests for the first time or there is no historical version data, directly send the full data;

[0018] Compress the incremental data using a dictionary-based compression algorithm, and the compression dictionary is dynamically generated according to historical high-frequency data fields;

[0019] Send the compressed incremental data to the client.

[0020] Furthermore, the MD5 check step also includes:

[0021] After the client merges the incremental data, it generates a local MD5 value and compares it with the pre-calculated MD5 value on the server side;

[0022] If the check fails three times in a row, trigger the client to force a refresh and re-establish a long connection.

[0023] Furthermore, the server's prediction of the user's next operation behavior based on the LSTM time series model specifically includes:

[0024] The server predicts the pages that may be accessed or the functions that may be triggered within a preset future time period based on the user's current page and operation path according to the LSTM time series model;

[0025] When the prediction confidence is lower than the preset threshold, suspend the pre-downstream of data and switch to the on-demand request mode.

[0026] Further, the version number maintenance uses a globally unique incremental identifier, and the version number is forced to be updated each time the data changes.

[0027] Further, the real-time data update method based on AI prediction further includes a step of dynamically adjusting the network transmission strategy, specifically including:

[0028] Automatically switch the incremental compression algorithm or degrade to full-volume transmission according to the client network latency or bandwidth fluctuation;

[0029] When a data merge conflict is detected, adopt the timestamp priority strategy to overwrite the local data or trigger manual intervention.

[0030] The present invention also provides a real-time data update device based on AI prediction, including:

[0031] A historical data collection module for the server to collect the historical behavior data set of users, and the historical behavior data includes the user's access frequency, time period preference, browsing data, geographical features, age information, the APP page currently accessed by the user, the button clicked on the user's current page, the user page operation path and events;

[0032] A model training module for training the historical behavior data set using the LSTM time series model, dividing the historical behavior data set into a training set and a test set according to a ratio, and optimizing the LSTM time series model by adjusting the hidden layer dimension, the number of LSTM layers or the learning rate parameter until the prediction accuracy of the test set exceeds 70%;

[0033] A long connection establishment module for the client to establish a long connection with the server and report the user's current operation behavior to the server in real time;

[0034] An operation behavior prediction module for the server to predict the user's next operation behavior based on the LSTM time series model and prepare the page or function data corresponding to the predicted operation in advance;

[0035] An incremental data calculation module for the server to compare the local version number carried by the client with the server version number by maintaining the version number of the data item, calculate the incremental data using the Diff algorithm, and issue the full-volume data if the amount of incremental data exceeds the full-volume data threshold;

[0036] An incremental data compression module for compressing the incremental data using a compression algorithm and actively sending it to the client through the long connection;

[0037] A full-volume data update module for the client to perform MD5 verification after merging the incremental data, and request a full-volume data update if the verification is inconsistent.

[0038] The beneficial effects of the present invention are as follows: By predicting user behavior, the waiting time for page refreshing is reduced; incremental update and data compression technologies reduce unnecessary data transmission; and the optimized data distribution strategy can improve data transmission speed; the use of conflict resolution and error recovery mechanisms ensures the accuracy of data transmission. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] The specific structure of the present invention will be described in detail below with reference to the accompanying drawings.

[0040] Figure 1 It is a flowchart of a real-time data update method based on AI prediction according to an embodiment of the present invention;

[0041] Figure 2 It is a flowchart of predicting the next operation behavior of a user according to an embodiment of the present invention;

[0042] Figure 3 It is a flowchart of massive data distribution according to an embodiment of the present invention;

[0043] Figure 4 It is a block diagram of a real-time data update device based on AI prediction according to an embodiment of the present invention;

[0044] Figure 5 It is a schematic block diagram of a computer device according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0045] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0046] It should be understood that when used in this specification and the appended claims, the terms "comprises" and "comprising" indicate the presence of the described features, wholes, steps, operations, elements, and / or components, but do not preclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their combinations.

[0047] It should also be understood that the terms used in this specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in this specification of the present invention and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an", and "the" are intended to include the plural forms.

[0048] It should also be further understood that the term "and / or" used in the specification and appended claims of the present invention refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0049] As Figure 1 shown, an embodiment of the present invention is: a real-time data update method based on AI prediction, including the following steps:

[0050] S10. The server collects the historical behavior data set of the user, and the historical behavior data includes the user's access frequency, time period preference, browsing data, geographical features, age information, the APP page currently accessed by the user, the button clicked on the user's current page, the user's page operation path, and events.

[0051] In this embodiment, the historical behavior data set is collected through a buried point SDK, specifically including: for example, the number of times the user accesses the product details page per hour in the e-commerce APP, the residence time on the shopping cart page from 19:00 to 21:00 every day, the click heat map of home appliance categories by users in a certain region, and the operation path sequence of the user from the home page → search bar → product list → order placement page. The data is stored using a sharded HBase cluster and horizontally sharded according to the hash value of the user ID. This embodiment can achieve structured storage of multi-dimensional user behavior characteristics, support high-concurrency real-time data writing, and provide a high-quality data foundation for model training.

[0052] S20. Use the LSTM time series model to train the historical behavior data set. The historical behavior data set is divided into a training set and a test set according to a ratio, and the LSTM time series model is optimized by adjusting the hidden layer dimension, the number of LSTM layers, or the learning rate parameter until the prediction accuracy of the test set exceeds 70%.

[0053] In this embodiment, the LSTM time series model adopts a three-layer stacked structure, the hidden layer dimension is set to 128, and the learning rate is adjusted to 0.001 when using the Adam optimizer. During training, the sliding window method is used, and the user's continuous 10 operations are used as the input sequence to predict the 11th operation type. When the accuracy of the test set reaches 75%, the model parameters are frozen, and the model is quantized and accelerated through TensorRT.

[0054] In a specific embodiment, the training of the LSTM time series model further includes:

[0055] Regularly add the newly generated behavior data of the user to the training set for dynamic optimization of the LSTM time series model;

[0056] Train through the LSTM time series model built into the PyTorch tool, and only adjust a single parameter each time to optimize the model performance.

[0057] S30. The client establishes a long connection with the server and reports the user's current operation behavior to the server in real time.

[0058] In this embodiment, the client establishes a long connection through the MQTT protocol and reports the device fingerprint (including screen resolution, OS version), click coordinates, and operation timestamp in real time when the user clicks the "Add to Cart" button. The server uses the Kafka message queue for behavior event stream processing.

[0059] S40. The server predicts the user's next operation behavior based on the LSTM time series model and prepares the page or function data corresponding to the predicted operation in advance.

[0060] Among them, as Figure 2 shown, step S40 specifically includes:

[0061] S41. The server predicts the pages that may be accessed or the functions that may be triggered within a preset future time period based on the LSTM time series model according to the user's current page and operation path;

[0062] S42. When the prediction confidence is lower than the preset threshold, suspend the pre-downstream of data and switch to the on-demand request mode.

[0063] In this embodiment, when it is detected that the user frequently browses the mobile phone category at 20:15 Beijing time and the stay duration is greater than 30 seconds, the model predicts that there is an 83% probability of entering the price comparison page. The server preloads the price comparison data of different merchants of the same brand and generates an incremental package.

[0064] S50. The server compares the local version number carried by the client with the server version number by maintaining the version number of the data item, calculates the incremental data using the Diff algorithm, and issues the full data if the amount of incremental data exceeds the full data threshold.

[0065] Among them, the version number maintenance uses a globally unique incremental identifier, and the version number is forcibly updated after each data change.

[0066] In this embodiment, the version number adopts the structure of "data table name_Unix timestamp_hash suffix".

[0067] S60. Compress the incremental data using a compression algorithm and actively issue it to the client through the long connection.

[0068] Among them, as Figure 3 shown, step S60 specifically includes:

[0069] S61. If the client requests for the first time or there is no historical version data, directly issue the full data;

[0070] S62. Compress the incremental data using a dictionary-based compression algorithm, and the compression dictionary is dynamically generated based on historical high-frequency data fields;

[0071] S63. Send the compressed incremental data to the client.

[0072] In this embodiment, the LZ78 dictionary compression algorithm is adopted to pre-generate a dictionary file containing high-frequency fields. When the incremental data contains 20 price records, the compression rate is increased from 60% to 85%, and it is transmitted in fragments through the QUIC protocol in a long connection channel.

[0073] S70. After the client merges the incremental data, perform an MD5 check. If the check is inconsistent, request a full data update.

[0074] In a specific embodiment, the MD5 check step further includes:

[0075] After the client merges the incremental data, generate a local MD5 value and compare it with the MD5 value pre-calculated by the server;

[0076] If the check fails three times in a row, trigger the client to forcibly refresh and re-establish a long connection.

[0077] In this embodiment, after the client merges the data, generate an MD5 value. If it matches the MD5 value sent by the server, the update is successful. If the check code does not match three times in a row, trigger the connection reconstruction process.

[0078] In a specific embodiment, the real-time data update method based on AI prediction further includes a step of dynamically adjusting the network transmission strategy, specifically including:

[0079] Automatically switch the incremental compression algorithm or degrade to full-volume transmission according to the client network latency or bandwidth fluctuation;

[0080] When a data merge conflict is detected, adopt the timestamp priority strategy to overwrite the local data or trigger manual intervention.

[0081] In this embodiment, when it is detected that the client network RTT is greater than 300 ms, automatically switch to Brotl i-3 level compression; when the bandwidth is less than 1 Mbps, degrade to full-volume transmission. When there is a data conflict, preferentially use the server timestamp to overwrite the client timestamp data.

[0082] In summary, by adopting the technical solution of this application, through AI-driven automated data updates, the database update efficiency can be increased by more than 80%, and manual intervention is reduced by 90%; achieve millisecond-level response through the LSTM time series model; add a feedback adjustment mechanism to improve the prediction accuracy; adopt the Diff algorithm and version control technology to reduce the data conflict rate by 95% and support strong consistency in high-concurrency scenarios.

[0083] As Figure 4 shown, an embodiment of the present invention further provides a real-time data update device based on AI prediction, including:

[0084] A historical data collection module 10, configured to collect a set of historical behavior data of a user by a server, where the historical behavior data includes the user's access frequency, time period preference, browsing data, geographical features, age information, the APP page currently accessed by the user, the button clicked on the user's current page, the user's page operation path, and events;

[0085] A model training module 20, configured to train the set of historical behavior data using an LSTM time series model, divide the set of historical behavior data into a training set and a test set according to a ratio, and optimize the LSTM time series model by adjusting the hidden layer dimension, the number of LSTM layers, or the learning rate parameter until the prediction accuracy of the test set exceeds 70%;

[0086] A long connection establishment module 30, configured to establish a long connection between a client and a server, and report the user's current operation behavior to the server in real time;

[0087] An operation behavior prediction module 40, configured to predict the user's next operation behavior by the server based on the LSTM time series model, and prepare the page or function data corresponding to the predicted operation in advance;

[0088] An incremental data calculation module 50, configured to calculate incremental data by the server by maintaining the version number of data items, comparing the local version number carried by the client with the server version number, and using the Diff algorithm. If the amount of incremental data exceeds the full data threshold, the full data is sent down;

[0089] An incremental data compression module 60, configured to compress the incremental data using a compression algorithm, and actively send it to the client through a long connection;

[0090] A full data update module 70, configured to perform an MD5 check after the client merges the incremental data. If the check is inconsistent, a full data update is requested.

[0091] It should be noted that those skilled in the art can clearly understand that the specific implementation process of the above real-time data update device based on AI prediction can refer to the corresponding description in the foregoing method embodiment. For the convenience and conciseness of description, it will not be elaborated here.

[0092] The above real-time data update device based on AI prediction can be implemented in the form of a computer program, and the computer program can run on a computer device as Figure 5 shown.

[0093] Please refer to Figure 5 ,Figure 5 FIG. Figure 5 is a schematic block diagram of a computer device provided by an embodiment of the present application. The computer device 500 may be a terminal or a server. Among them, the terminal may be an electronic device with communication functions such as a smart phone, a tablet computer, a notebook computer, a desktop computer, a personal digital assistant, and a wearable device. The server may be an independent server or a server cluster composed of multiple servers.

[0094] Referring to Figure 5 , the computer device 500 includes a processor 502, a memory, and a network interface 505 connected through a system bus 501. Among them, the memory may include a non-volatile storage medium 503 and an internal memory 504.

[0095] The non-volatile storage medium 503 may store an operating system 5031 and a computer program 5032. The computer program 5032 includes program instructions. When the program instructions are executed, the processor 502 may be caused to execute a real-time data update method based on AI prediction.

[0096] The processor 502 is used to provide computing and control capabilities to support the operation of the entire computer device 500.

[0097] The internal memory 504 provides an environment for the operation of the computer program 5032 in the non-volatile storage medium 503. When the computer program 5032 is executed by the processor 502, the processor 502 may be caused to execute a real-time data update method based on AI prediction.

[0098] The network interface 505 is used for network communication with other devices. Those skilled in the art can understand that Figure 5 the structure shown in FIG. is only a block diagram of some structures related to the solution of the present application, and does not constitute a limitation on the computer device 500 to which the solution of the present application is applied. The specific computer device 500 may include more or fewer components than those shown in the figure, or combine some components, or have different component arrangements.

[0099] Among them, the processor 502 is used to run the computer program 5032 stored in the memory to implement the real-time data update method based on AI prediction as described above.

[0100] It should be understood that in the embodiments of the present application, the processor 502 may be a central processing unit (CPU), and the processor 502 may also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0101] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program includes program instructions, and the computer program can be stored in a storage medium, which is a computer-readable storage medium. The program instructions are executed by at least one processor in the computer system to implement the process steps of the embodiments of the above methods.

[0102] Therefore, the present invention also provides a storage medium. The storage medium may be a computer-readable storage medium. The storage medium stores a computer program, where the computer program includes program instructions. When the program instructions are executed by a processor, the processor executes the real-time data update method based on AI prediction as described above.

[0103] The storage medium may be a USB flash drive, a mobile hard disk, a read-only memory (ROM), a magnetic disk, an optical disc, or other computer-readable storage media that can store program codes.

[0104] Those of ordinary skill in the art can realize that the units and algorithm steps of the examples described in combination with the embodiments disclosed herein can be implemented by electronic hardware, computer software, or a combination of the two. To clearly illustrate the interchangeability of hardware and software, the components and steps of the examples have been generally described according to their functions in the above description. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present invention.

[0105] In several embodiments provided by the present invention, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of each unit is only a logical function division, and there may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed.

[0106] The steps in the method embodiments of the present invention can be adjusted, combined, and deleted according to actual needs. The units in the device embodiments of the present invention can be combined, divided, and deleted according to actual needs. In addition, the functional units in each embodiment of the present invention can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit.

[0107] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a terminal, or a network device, etc.) to execute all or part of the steps of the methods described in each embodiment of the present invention.

[0108] As described above, the above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of various equivalent modifications or substitutions, and these modifications or substitutions should all be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims

1. A real-time data update method based on AI prediction, characterized in that: The following steps are involved: The server collects a set of historical behavior data of the user, which includes the user's access frequency, time period preference, browsing data, regional characteristics, age information, the APP page currently visited by the user, the button clicked by the user on the current page, the user's page operation path and events; The historical behavior data set is trained using an LSTM time series model, the historical behavior data set is divided into a training set and a test set in proportion, and the LSTM time series model is optimized by adjusting the hidden layer dimension, the number of LSTM layers or the learning rate parameter until the prediction accuracy of the test set exceeds 70%; The client establishes a persistent connection with the server and reports the user's current operation behavior to the server in real time; The server predicts the user's next operation behavior based on the LSTM time series model, and prepares the page or function data corresponding to the predicted operation in advance; The server maintains the version number of the data item, compares the local version number carried by the client with the server version number, and uses the Diff algorithm to calculate the incremental data. If the incremental data volume exceeds the full data threshold, the full data is sent; The incremental data is compressed using a compression algorithm and actively sent to the client through a long connection; After merging the incremental data, the client performs an MD5 check. If the check is inconsistent, it requests a full data update.

2. The real-time data updating method based on AI prediction according to claim 1 is characterized in that: The LSTM time series model training also includes: Regularly adding new user behavior data to the training set to dynamically optimize the LSTM time series model; The built-in LSTM time series model of the PyTorch tool is used for training, and only a single parameter is adjusted each time to optimize the model performance.

3. The real-time data updating method based on AI prediction according to claim 1 is characterized in that: The step of compressing the incremental data using a compression algorithm and actively sending it to the client through a long connection specifically includes: If the client requests for the first time or there is no historical version data, the full data will be sent directly; The incremental data is compressed based on the dictionary-based compression algorithm, and the compression dictionary is dynamically generated based on the historical high-frequency data fields; Send the compressed incremental data to the client.

4. The real-time data updating method based on AI prediction according to claim 1 is characterized in that: The MD5 verification step also includes: The client merges the incremental data to generate a local MD5 value and compares it with the MD5 value pre-calculated by the server; If the verification fails three times in a row, the client is triggered to force a refresh and re-establish a persistent connection.

5. The real-time data updating method based on AI prediction according to claim 1 is characterized in that: The server predicts the user's next operation behavior based on the LSTM time series model, specifically including: The server predicts the pages that may be visited or the functions that may be triggered within a preset time period in the future based on the LSTM time series model according to the user's current page and operation path; When the prediction confidence is lower than the preset threshold, data pre-delivery is suspended and switched to on-demand request mode.

6. The real-time data updating method based on AI prediction according to claim 1 is characterized in that: The version number maintenance adopts a globally unique incremental identifier, and the version number is forcibly updated after each data change.

7. The real-time data updating method based on AI prediction according to claim 1 is characterized in that: It also includes the strategy steps for dynamically adjusting network transmission, including: Automatically switch to incremental compression algorithm or downgrade to full transmission based on client network latency or bandwidth fluctuations; When a data merge conflict is detected, a timestamp priority strategy is used to overwrite local data or trigger manual intervention.

8. A real-time data updating device based on AI prediction, characterized in that: include: The historical data collection module is used for the server to collect the user's historical behavior data set, which includes the user's access frequency, time period preference, browsing data, regional characteristics, age information, the APP page currently visited by the user, the button clicked by the user on the current page, the user's page operation path and events; A model training module, for training the historical behavior data set using an LSTM time series model, wherein the historical behavior data set is divided into a training set and a test set in proportion, and the LSTM time series model is optimized by adjusting the hidden layer dimension, the number of LSTM layers or the learning rate parameter until the prediction accuracy of the test set exceeds 70%; The persistent connection establishment module is used to establish a persistent connection between the client and the server, and report the user's current operation behavior to the server in real time; An operation behavior prediction module is used for the server to predict the user's next operation behavior based on the LSTM time series model, and prepare the page or function data corresponding to the predicted operation in advance; The incremental data calculation module is used by the server to maintain the version number of the data item, compare the local version number carried by the client with the server version number, and use the Diff algorithm to calculate the incremental data. If the incremental data volume exceeds the full data threshold, the full data is sent; An incremental data compression module is used to compress the incremental data using a compression algorithm and actively send it to the client through a long connection; The full data update module is used by the client to perform MD5 verification after merging incremental data. If the verification is inconsistent, a full data update is requested.

9. A computer device, characterized in that: The computer device includes a memory and a processor, the memory stores a computer program, and the processor implements the real-time data update method based on AI prediction as described in any one of claims 1 to 7 when executing the computer program.

10. A storage medium, characterized in that: The storage medium stores a computer program, and when the computer program is executed by the processor, it can implement the real-time data updating method based on AI prediction as described in any one of claims 1 to 7.

Images