WO2023272858A1

WO2023272858A1 - Page resource caching method and apparatus, device, and medium

Info

Publication number: WO2023272858A1
Application number: PCT/CN2021/109049
Authority: WO
Inventors: 余鸿飞
Original assignee: 未鲲(上海)科技服务有限公司
Priority date: 2021-06-28
Filing date: 2021-07-28
Publication date: 2023-01-05
Also published as: CN113419784A

Abstract

A page resource caching method and apparatus, a device, and a medium, relating to the technical field of artificial intelligence. The method comprises: obtaining a page loading completion signal for an i-th accessed page, the page loading completion signal carrying a user identifier to be predicted and an i-th accessed page address real value (S1); in response to the page loading completion signal, obtaining a target user portrait according to the user identifier to be predicted (S2); using a first preset page prediction model to predict a (i+1)-th accessed page address according to the i-th accessed page address real value and the target user portrait, to obtain a (i+1)-th accessed page address predicted value (S3); searching in a local cache according to the (i+1)-th accessed page address predicted value (S4); when a cache result is not found, obtaining a page resource from a server according to the (i+1)-th accessed page address predicted value, to obtain a page resource to be cached (S5); and storing the page resource to be cached in the local cache (S6). Therefore, a page resource that is likely to be accessed next time by a user is automatically cached to a local cache on the premise of no perception by the user, thereby increasing the speed of loading of an unopened page.

Description

Cache method, device, equipment and medium for page resources

This application claims the priority of the Chinese patent application with the application number 202110722084.X submitted to the China Patent Office on June 28, 2021, and the title of the invention is "page resource caching method, device, equipment and medium", the entire content of which is passed References are incorporated in this application.

technical field

The present application relates to the technical field of artificial intelligence, in particular to a method, device, device and medium for caching page resources.

Background technique

When a user visits a page, the client will first download the static resources of the page, such as CSS (Cascading Style Sheet), JS (JavaScript), HTML (Hypertext Markup Language), image files, etc. All the static resources of the page When the download is complete, the entire page loads normally. This loading method has the following problems: If the static resources of the page are relatively large, or the user's current network status is not ideal, it will take more time to download the static resources. When the user is waiting for the page to load, he often sees a long time The loading animation or blank screen of the page affects the user experience. The inventor realizes that in order to solve the problem of slow loading, the client uses the built-in caching function. When the user accesses the page, the client will cache the static resources of the page locally. When the user opens the page again, the client directly downloads the The static resource files are read in the local cache, thus improving the loading speed. However, the method of using the cache function of the client to cache the static resources of the visited pages depends on the user having opened the page in advance, and the caching strategy is invalid for the pages that the user has not opened, so that the loading of the pages that have not been opened cannot be improved. speed.

technical problem

The invention aims to solve the technical problem that the prior art method of caching the static resources of the visited pages by using the built-in caching function of the client cannot improve the loading speed of the unopened pages.

technical solution

The main purpose of this application is to provide a caching method, device, equipment and medium for page resources, aiming to solve the problem that the prior art uses the client's own caching function to cache the static resources of pages that have been visited, and cannot improve the performance of pages that have not been opened. Technical issues with page load speed.

In order to achieve the purpose of the above invention, this application proposes a method for caching page resources, the method comprising:

Acquiring the page loading completion signal of the i-th visit page, the page loading completion signal carrying the user identifier to be predicted and the real value of the i-th visit page address;

Responding to the page loading completion signal, acquiring a target user portrait according to the user identifier to be predicted;

Using the first preset page prediction model, predicting the page address of the i+1th visit according to the real value of the page address of the ith visit and the target user portrait, to obtain the predicted value of the page address of the i+1th visit;

Searching the predicted value of the i+1th access page address in the local cache;

When the cache result is not found, obtain the page resource from the server according to the i+1th access page address prediction value, and obtain the page resource to be cached;

storing the page resources to be cached in the local cache.

The present application also proposes a caching device for page resources, and the device includes:

The page loading completion signal acquisition module is used to obtain the page loading completion signal of the i-th visit page, and the page loading completion signal carries the user identifier to be predicted and the real value of the i-th visit page address;

A target user portrait acquisition module, configured to acquire a target user portrait according to the user identifier to be predicted in response to the page loading completion signal;

The page address prediction module is configured to use the first preset page prediction model to predict the page address of the i+1th visit according to the real value of the page address of the ith visit and the target user portrait, to obtain the i+1th visit The predicted value of the page address of the second visit;

A cache lookup module, configured to look up the i+1th access page address prediction value in a local cache;

The page resource determination module to be cached is used to obtain the page resource from the server according to the i+1th access page address prediction value to obtain the page resource to be cached when no cached result is found;

A storage module, configured to store the page resource to be cached in the local cache.

The present application also proposes a computer device, including a memory and a processor, the memory stores a computer program, and the processor implements the following method steps when executing the computer program:

storing the page resources to be cached in the local cache.

The present application also proposes a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, the following method steps are implemented:

storing the page resources to be cached in the local cache.

Beneficial effect

The page resource caching method, device, device and medium of the present application obtain the page loading completion signal of the i-th visit page first, and the page loading completion signal carries the user identifier to be predicted and the real value of the address of the i-th visit page , secondly in response to the page loading completion signal, obtain the target user portrait according to the user identifier to be predicted, adopt the first preset page prediction model, and perform the process according to the real value of the ith visit page address and the target user portrait Predict the page address of the i+1th access, obtain the predicted value of the i+1th accessed page address, and then search the predicted value of the i+1th accessed page address in the local cache, when no cached result is found , obtain page resources from the server according to the predicted value of the i+1th access page address, obtain the page resources to be cached, and finally store the page resources to be cached in the local cache, so that the user does not perceive automatically cache the page resource that the user may visit next time to the local cache, and when the user actually visits the predicted page address, the page resource will be obtained from the local cache for loading, thereby reducing the page loading time and improving the unopened The loading speed of the past pages improves the user experience.

Description of drawings

FIG. 1 is a schematic flowchart of a method for caching page resources according to an embodiment of the present application;

FIG. 2 is a schematic block diagram of a cache device for page resources according to an embodiment of the present application;

FIG. 3 is a schematic block diagram of a computer device according to an embodiment of the present application.

The realization, functional features and advantages of the present application will be further described in conjunction with the embodiments and with reference to the accompanying drawings.

Embodiments of the present invention

In order to make the purpose, technical solution and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described here are only used to explain the present application, and are not intended to limit the present application.

Referring to FIG. 1, an embodiment of the present application provides a method for caching page resources, the method comprising:

S1: Obtain the page loading completion signal of the i-th visited page, the page loading completion signal carries the user identifier to be predicted and the real value of the i-th visited page address;

S2: Responding to the page loading completion signal, acquiring a target user portrait according to the user identifier to be predicted;

S3: Using the first preset page prediction model, predicting the page address of the i+1th visit according to the real value of the page address of the ith visit and the target user portrait, and obtaining the page address prediction of the i+1th visit value;

S4: Search the predicted value of the i+1th access page address in the local cache;

S5: When the cached result is not found, obtain the page resource from the server according to the predicted value of the i+1th access page address, and obtain the page resource to be cached;

S6: Store the page resource to be cached in the local cache.

In this embodiment, by first obtaining the page loading completion signal of the i-th access page, the page loading completion signal carries the user identifier to be predicted and the real value of the i-th access page address, and secondly responding to the page loading completion signal, according to the Describe the user ID to be predicted to obtain the target user portrait, adopt the first preset page prediction model, and predict the page address of the i+1 visit according to the real value of the page address of the ith visit and the target user portrait, and obtain The i+1th access page address prediction value, and then look up the i+1th access page address prediction value in the local cache, and when no cached result is found, according to the i+1th access page The address prediction value obtains the page resource from the server, obtains the page resource to be cached, and finally stores the page resource to be cached in the local cache, so that the page that the user may visit next time is automatically stored without the user's perception The resources are cached in the local cache. When the user actually visits the predicted page address, the page resources will be obtained from the local cache for loading, thereby reducing the page loading time, increasing the loading speed of unopened pages, and improving the user experience.

For S1, the client loads the i-th visit page of the user ID to be predicted, and generates a page loading completion signal when the i-th visit page is loaded, and uses the page address of the i-th visit page as the i-th visit page address For the real value, the user identifier to be predicted and the real value of the address of the i-th visited page are used as parameters carried by the page loading completion signal.

The page loading completion signal is the signal generated when the i-th access page is loaded.

The client can be a mobile device client, a computer client, or a browser.

Optionally, the i-th visited page refers to the page currently browsed by the user. A page is a Web (Global Wide Area Network) page.

The user ID to be predicted is the user ID of the user who visits the i-th visit page. The user identifier may be data that uniquely identifies a user, such as a user name and a user ID.

The real value of the address of the i-th visit page, that is, the URL (Uniform Resource Locator) address of the page that the user actually visits the i-th time corresponding to the user ID to be predicted.

For S2, when the page loading completion signal is received, in response to the page loading completion signal, acquire the user portrait corresponding to the user identifier to be predicted as the target user portrait.

Optionally, when the page loading completion signal is received, the page prediction interface is called in response to the page loading completion signal, and the user portrait corresponding to the user identifier to be predicted is obtained as the target user portrait, so that the user portrait and Client decoupling helps to improve the stability of client services.

Optionally, the step of acquiring the user profile corresponding to the user ID to be predicted as the target user profile includes: acquiring a user profile library; searching the user ID to be predicted in the user profile library, The user portrait corresponding to the user identification found in the user portrait database is used as the target user portrait.

The user portrait library includes: user ID and user portrait, and each user ID corresponds to a user portrait.

User portraits include but are not limited to: gender, age, education, location, device information, user classification tags, and historical browsing records.

When this application is applied to the financial industry, the user profile also includes: user risk level, user security level, and historical investment records (product type, amount).

For S3, input the real value of the page address of the i-th visit and the target user portrait into the first preset page prediction model to predict the page address of the i+1-th visit, and use the predicted page address as the first page address i+1 access page address prediction value.

Optionally, calling the page prediction interface, inputting the real value of the page address of the i-th visit and the target user portrait into the first preset page prediction model to predict the page address of the i+1-th visit, The predicted page address is used as the predicted value of the i+1th visit page address, thereby decoupling the first preset page prediction model from the client, which is beneficial to improving the stability of the client service.

For example, the real value of the page address of the i-th visit is the real value of the page address of the page visited by the user m for the i-th time, then the predicted value of the page address for the i+1 visit is to predict that the user m will visit the page for the i+1 time The page address of the accessed page is not specifically limited in this example.

The target user portrait is used as the input of the first preset page prediction model because users with the same or similar portraits may enter similar pages next time. For example, users with high risk tolerance will tend to choose fund products with greater volatility when browsing the investment list page, so before the user enters the fund product page, the corresponding fund products with greater volatility will be selected in advance. The page resources of the page are cached in the client's local cache, which is not limited in this example.

Wherein, the first preset page prediction model is a model trained based on an AI (artificial intelligence) deep learning model.

For S4, search in the client's local cache according to the i+1th access page address prediction value, when the page resource corresponding to the i+1th access page address prediction value is found in the client's local cache When determining that the cache lookup result is cached, when the page resource corresponding to the i+1th access page address prediction value is not found in the local cache of the client, it is determined that the cache lookup result is not cached.

Page resources, that is, static resources of the page, such as CSS (Cascading Style Sheet), JS (JavaScript), HTML (Hypertext Markup Language), image files, etc.

For S5, when the cache result is not found, it means that the page resource corresponding to the i+1th access page address prediction value is not stored in the local cache of the client, so according to the i+1th access page address prediction The value obtains page resources from the server, and uses all obtained page resources as page resources to be cached.

For S6, the page resource to be cached is stored in the local cache of the client, so that when the user accesses the page corresponding to the predicted value of the i+1th access page address, the page resource is obtained from the local cache of the client Loading, thereby reducing the page loading time, improving the loading speed of pages that have not been opened, and improving user experience.

In one embodiment, the above-mentioned step of acquiring a target user portrait according to the user identifier to be predicted includes:

S21: Call the page prediction interface, input the user identification to be predicted into the user portrait model, and obtain the target user portrait output by the user portrait model, wherein the user portrait model obtains the target user portrait according to the user identification to be predicted The user data of the portrait, the user portrait is performed according to the user data to be portraited, and the target user portrait is obtained.

In this embodiment, by calling the page prediction interface, the user identification to be predicted is input to the user portrait model, and the target user portrait output by the user portrait model is obtained, and the user portrait model is based on the user identification to be predicted Acquiring the user data to be profiled, and performing user portraits according to the user data to be profiled, to obtain the target user portrait, which is conducive to real-time generation of user portraits, improving the accuracy of the determined target user portraits, thereby improving the next time The accuracy of the visited page address prediction; decoupling the user portrait from the client through the page prediction interface helps to improve the stability of the client service.

For S21, call the page prediction interface, input the user ID to be predicted into the user portrait model, and the user portrait model will obtain the user data to be profiled when receiving the user ID to be predicted, according to the The user data to be profiled is profiled in real time, and the data obtained from the profile is used as the profile of the target user.

The user data to be profiled includes, but is not limited to: basic user information, user classification tags, and historical browsing records. Basic user information includes, but is not limited to: gender, age, education, location, and device information. The user classification label is a category label for classifying users. Browsing history records are records of users' historical browsing pages.

The user portrait model is a model obtained based on neural network training.

In one embodiment, the above step of searching the predicted value of the i+1th access page address in the local cache includes:

S41: Look up the predicted value of the i+1th access page address in the local cache;

S42: When no page address is found, determine that no cached result is found in the local cache;

S43: When the page address is found, obtain the version identifier from the server according to the predicted value of the i+1th access page address, obtain the version identifier to be cached, and store the i+1th access in the local cache The version identifier of the page resource corresponding to the predicted value of the page address accessed once is used as the local cache version identifier;

S44: Compare the version identifier to be cached with the locally cached version identifier;

S45: When the version identifier to be cached is the same as the local cache version identifier, determine that a cached result is found in the local cache;

S46: When the version identifier to be cached is different from the local cache version identifier, determine that no cached result is found in the local cache.

In this embodiment, the version identifier to be cached is compared with the locally cached version identifier, and when the version identifier to be cached is the same as the locally cached version identifier, it is determined that the cache lookup result is cached , when the version identifier to be cached is different from the locally cached version identifier, determine that the cached lookup result is not cached, thereby improving the accuracy of the cached lookup result.

For S42, when the page address is not found, it means that the i+1th access page address prediction value does not cache page resources in the local cache of the client, so it can be determined that no cached result is found in the local cache .

For S43, when the page address is found, it means that the i+1th access page address prediction value has page resources cached in the local cache of the client, and the i+1th access page address prediction value is obtained from the The server obtains the version identifier, and uses the obtained version identifier as the version identifier to be cached; uses the version identifier of the page resource corresponding to the i+1th access page address prediction value in the client's local cache as the local cache version identifier . That is to say, the version identifier to be cached is the version identifier of the latest version of the page resource on the server, and the locally cached version identifier is the version identifier of the locally cached page resource.

For S45, when the version identifier to be cached is the same as the local cached version identifier, it means that the latest version of the page resource on the server side is the same as the page resource in the local cache, so it is determined that the cached result is found in the local cache .

For S46, when the version identifier to be cached is different from the local cache version identifier, it means that the latest version of the page resource on the server side is different from the page resource in the local cache, so it is determined that no to the cached result.

In one embodiment, the above step of obtaining page resources from the server according to the predicted value of the i+1th access page address, and obtaining the page resources to be cached includes:

S51: Generate a page resource acquisition request according to the predicted value of the i+1th access page address, and send the page resource acquisition request to the server;

S52: Obtain the page resource sent by the server according to the page resource acquisition request as the page resource to be cached.

In this embodiment, a page resource acquisition request is generated according to the predicted value of the i+1th access page address, and then the page resource sent by the server according to the page resource acquisition request is acquired as the page resource to be cached, In this way, the latest version of the page resource on the server is obtained, which provides a basis for caching the page resource in the local cache of the client in advance.

For S51, generate a page resource acquisition request according to the predicted value of the i+1th access page address, that is, when generating the page resource acquisition request, encapsulate the i+1th accessed page address prediction value into the page Parameters for resource fetch requests.

For S52, when the server receives the page resource acquisition request, it first parses the predicted value of the i+1th access page address from the page resource acquisition request, and then parses the parsed The predicted value of the page address for i+1 visits is searched from the page resource library, and the page resource corresponding to the page address found in the page resource library is sent to the client corresponding to the page resource acquisition request.

The page resource library includes: page addresses and page resources, and each page address corresponds to a page resource.

In one embodiment, after the above-mentioned step of obtaining the page loading completion signal of the i-th visited page, it further includes:

S71: Obtain the real value of the page address of the i-1th visited page according to the user identifier to be predicted and the real value of the i-th visited page address, and obtain the real value of the i-1th visited page address;

S72: Obtain a user portrait to be predicted according to the user identifier to be predicted;

S73: Train a second preset page prediction model according to the actual value of the i-1th accessed page address, the actual value of the i-th accessed page address, and the user portrait to be predicted;

S74: Update the first preset page prediction model according to the trained second preset page prediction model.

In this embodiment, when the page loading completion signal of the i-th accessed page is acquired, the second preset page prediction model is trained, and the first preset is updated according to the trained second preset page prediction model. The page prediction model eliminates the need to manually collect training data, and quickly improves the accuracy of the page address prediction for the next visit of the first preset page prediction model; by separating the trained model from the service-providing model, it is beneficial to improve The response efficiency of the prediction model for the first preset page of the service is provided.

For S71, according to the user identifier to be predicted and the real value of the i-th visited page address, the real value of the page address of the i-1th accessed page is obtained from the local cache of the client, and the obtained i-th The actual value of the page address of the page accessed by 1 is used as the actual value of the i-1th accessed page address.

For S72, input the to-be-predicted user identifier into a user portrait model, and acquire the to-be-predicted user portrait sent by the user portrait model according to the to-be-predicted user identifier.

Optionally, calling the page training interface, inputting the user identity to be predicted into the user portrait model, and obtaining the user portrait to be predicted sent by the user portrait model according to the user identity to be predicted, so that Decoupling the user profile from the client will help improve the stability of the client service.

Optionally, the step of obtaining the user profile to be predicted according to the user ID to be predicted includes: acquiring the user profile library; searching the user ID to be predicted in the user profile library, The user portrait corresponding to the user identifier found in the user portrait database is used as the user portrait to be predicted.

For S73, input the actual value of the i-1th visited page address and the user portrait to be predicted into the second preset page prediction model to predict the page address of the i-th visit, according to the predicted page address and The actual value of the i-th accessed page address is used to train the second preset page prediction model.

Calling the page training interface, inputting the actual value of the i-1th visited page address and the user portrait to be predicted into the second preset page prediction model to predict the page address of the i-th visit, and obtain The page address of the second preset page prediction model and the actual value of the i-th visit page address are used to train the second preset page prediction model, thereby decoupling the second preset page prediction model from the client, which is beneficial to improving the stability of the client service.

Wherein, the model structures of the first preset page prediction model and the second preset page prediction model are the same.

For S74, update the model parameters of the first preset page prediction model according to the trained model parameters of the second preset page prediction model.

It can be understood that, in another embodiment, the first preset can also be made according to the real value of the address of the i-1th visited page, the real value of the address of the i-th visited page, and the user portrait to be predicted. The page prediction model is trained, which is not limited here.

In one embodiment, the above step of training the second preset page prediction model according to the actual value of the i-1th visited page address, the true value of the i-th visited page address and the user portrait to be predicted ,include:

S731: Use the real value of the address of the i-th accessed page as the calibration value of the i-th accessed page;

S732: Input the user portrait to be predicted and the actual value of the i-1th accessed page address into the second preset page prediction model to predict the page address of the i-th accessed page, and obtain the predicted value of the i-th accessed page address ;

S733: Train the second preset page prediction model according to the i-th accessed page address prediction value and the i-th accessed page calibration value.

In this embodiment, the user portrait to be predicted and the actual value of the i-1th visited page address are input into the second preset page prediction model to predict the page address of the i-th visit. According to the i-th visited page The address prediction value and the i-th visit page calibration value train the second preset page prediction model, thereby eliminating the need to manually collect training data, and quickly improving the performance of the first preset page prediction model for the next visit. Accuracy of Page Address Prediction.

For S731, since the real value of the i-th visited page address is the address of the page actually visited by the user, the real value of the i-th visited page address is used as the i-th visited page calibration value.

For S732, input the user portrait to be predicted and the actual value of the i-1th visited page address into the second preset page prediction model to predict the page address of the i-th visit, and use the predicted page address as Predicted address value of the i-th accessed page.

For S733, input the i-th accessed page address prediction value and the i-th accessed page calibration value into a cross-entropy loss function to calculate a loss value to obtain a target loss value, and update the second predicted value according to the target loss value Set the parameters of the page prediction model.

In one embodiment, the step of updating the first preset page prediction model according to the trained second preset page prediction model includes:

S741: Using the preset model parameter update time, update the model parameters in the first preset page prediction model according to the model parameters in the second preset page prediction model, wherein the first preset page prediction model The model has the same model structure as the second preset page prediction model.

In this embodiment, the model parameters in the first preset page prediction model are updated by using the preset model parameter update time, thereby avoiding that frequent updates affect the ability of the first preset page prediction model to provide services.

For S741, the preset time for updating model parameters includes but is not limited to: 3 o'clock every day.

Wherein, extracting model parameters from the second preset page prediction model to obtain a model parameter matrix to be updated; updating model parameters in the first preset page prediction model according to the model parameter matrix to be updated, The updated first preset page prediction model is used to predict the page address of the next visit, and only the model parameters are updated, thereby reducing the amount of updated data and improving the updating efficiency.

Referring to FIG. 2, the present application also proposes a caching device for page resources, and the device includes:

The page loading completion signal acquisition module 100 is used to obtain the page loading completion signal of the i-th visit page, and the page loading completion signal carries the user identifier to be predicted and the real value of the i-th visit page address;

Target user portrait acquisition module 200, configured to respond to the page loading completion signal, and acquire a target user portrait according to the user identifier to be predicted;

The page address prediction module 300 is configured to use the first preset page prediction model to predict the page address of the i+1th visit according to the real value of the page address of the ith visit and the target user portrait, and obtain the i+1th visit. 1 access page address prediction value;

A cache lookup module 400, configured to look up the i+1th access page address prediction value in a local cache;

The page resource determination module 500 to be cached is used to obtain the page resource from the server according to the i+1th access page address prediction value to obtain the page resource to be cached when no cached result is found;

The storage module 600 is configured to store the page resource to be cached in the local cache.

Referring to FIG. 3 , an embodiment of the present application also provides a computer device, which may be a server, and its internal structure may be as shown in FIG. 3 . The computer device includes a processor, memory, network interface and database connected by a system bus. Among them, the processor designed by the computer is used to provide calculation 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, computer programs and databases. The memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database of the computer device is used to store data such as the caching method of the page resource. 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, a page resource caching method is realized. The caching method of the page resource includes: obtaining the page loading completion signal of the i-th access page, the page loading completion signal carrying the user identifier to be predicted and the real value of the i-th access page address; responding to the completion of the page loading signal, obtaining a target user portrait according to the user identifier to be predicted; using the first preset page prediction model, performing the i+1th visit page according to the real value of the i-th visit page address and the target user portrait Address prediction, obtaining the i+1th access page address prediction value; searching the i+1th access page address prediction value in the local cache; when no cache result is found, according to the i+1th access page address prediction value The predicted value of the page address of the second visit obtains the page resource from the server to obtain the page resource to be cached; and stores the page resource to be cached in the local cache.

An embodiment of the present application also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, a method for caching page resources is implemented, including the steps of: obtaining the page load of the i-th accessed page Complete signal, the page loading completion signal carries the user identification to be predicted and the real value of the i-th visit page address; in response to the page loading completion signal, obtain the target user portrait according to the user identification to be predicted; adopt the first prediction A page prediction model is set, and the page address prediction of the i+1th visit is performed according to the real value of the page address of the ith visit and the target user portrait, and the predicted value of the page address of the i+1th visit is obtained; The predicted value of the page address of the i+1 visit is searched in the local cache; when the cache result is not found, the page resource is obtained from the server according to the predicted value of the page address of the i+1 visit, and the page resource to be cached is obtained. ; Store the page resource to be cached in the local cache.

The page resource caching method executed above firstly obtains the page loading completion signal of the i-th access page, the page loading completion signal carries the user identifier to be predicted and the real value of the i-th access page address, and secondly responds to the page Loading completion signal, obtaining target user portrait according to the user identifier to be predicted, using the first preset page prediction model, and performing the i+1th visit according to the real value of the i-th visit page address and the target user portrait The page address prediction of the i+1th access page address is obtained, and then the i+1th access page address prediction value is searched in the local cache. When the cache result is not found, according to the first The predicted value of the page address for i+1 visits obtains the page resource from the server, obtains the page resource to be cached, and finally stores the page resource to be cached in the local cache, thereby automatically saving the user The page resources that may be accessed next time are cached in the local cache. When the user actually visits the predicted page address, the page resources will be obtained from the local cache for loading, thereby reducing the page loading time and improving the loading speed of unopened pages , improving the user experience.

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 implemented through computer programs to instruct related hardware, and the computer programs can be stored in a non-volatile computer-readable memory In the medium, when the computer program is executed, it may include the processes of the embodiments of the above-mentioned methods. Wherein, any references to memory, storage, database or other media provided in the present application and used in the embodiments may include non-volatile and/or volatile memory. Nonvolatile 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 many forms such as Static RAM (SRAM), Dynamic RAM (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (SSRSDRAM), Enhanced SDRAM (ESDRAM), Synchronous Link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

The computer-readable storage medium may be non-volatile or volatile.

It should be noted that, in this document, the term "comprising", "comprising" or any other variation thereof is intended to cover a non-exclusive inclusion such that a process, apparatus, article or method comprising a set of elements includes not only those elements, It also includes other elements not expressly listed, or elements inherent in the process, apparatus, article, or method. Without further limitations, an element defined by the phrase "comprising a ..." does not preclude the presence of additional same elements in the process, apparatus, article or method comprising the element.

The above are only preferred embodiments of the application, and are not intended to limit the patent scope of the application. Any equivalent structure or equivalent process conversion made by using the specification and drawings of the application, or directly or indirectly used in other related All technical fields are equally included in the patent protection scope of the present application.

Claims

A method for caching page resources, wherein the method includes:

Acquiring the page loading completion signal of the i-th visit page, the page loading completion signal carrying the user identifier to be predicted and the real value of the i-th visit page address;

Responding to the page loading completion signal, acquiring a target user portrait according to the user identifier to be predicted;

Using the first preset page prediction model, predicting the page address of the i+1th visit according to the real value of the page address of the ith visit and the target user portrait, to obtain the predicted value of the page address of the i+1th visit;

Searching the predicted value of the i+1th access page address in the local cache;

When the cache result is not found, obtain the page resource from the server according to the i+1th access page address prediction value, and obtain the page resource to be cached;

storing the page resources to be cached in the local cache.
The method for caching page resources according to claim 1, wherein the step of acquiring a target user portrait according to the user identifier to be predicted comprises:

calling the page prediction interface, inputting the user identification to be predicted into the user portrait model, and obtaining the target user portrait output by the user portrait model, wherein the user portrait model obtains the user to be profiled according to the user identification to be predicted Data, perform user portrait according to the user data to be portraited, and obtain the target user portrait.
The method for caching page resources according to claim 1, wherein the step of searching the predicted value of the i+1th access page address in a local cache includes:

Searching the predicted value of the i+1th access page address in the local cache;

When the page address is not found, it is determined that no cached result is found in the local cache;

When the page address is found, obtain the version identifier from the server according to the i+1th access page address prediction value, obtain the version identifier to be cached, and store the i+1th access in the local cache The version identifier of the page resource corresponding to the predicted value of the accessed page address is used as the local cache version identifier;

comparing the version identifier to be cached with the locally cached version identifier;

When the version identifier to be cached is the same as the local cache version identifier, determine that a cached result is found in the local cache;

When the version identifier to be cached is different from the local cache version identifier, it is determined that no cached result is found in the local cache.
The method for caching page resources according to claim 1, wherein the step of obtaining page resources from the server according to the predicted value of the page address of the i+1 access to obtain the page resources to be cached includes:

Generate a page resource acquisition request according to the predicted value of the i+1th access page address, and send the page resource acquisition request to the server;

Acquire the page resource sent by the server according to the page resource acquisition request as the page resource to be cached.
The caching method for page resources according to claim 1, wherein, after the step of obtaining the page loading completion signal of the i-th accessed page, further comprising:

Obtain the real value of the page address of the i-1th visited page according to the user identifier to be predicted and the real value of the i-th visited page address, and obtain the real value of the i-1th visited page address;

Obtaining a user portrait to be predicted according to the user identifier to be predicted;

Training a second preset page prediction model according to the actual value of the i-1th visited page address, the true value of the i-th visited page address, and the user portrait to be predicted;

Updating the first preset page prediction model according to the trained second preset page prediction model.
The method for caching page resources according to claim 5, wherein, according to the actual value of the i-1th accessed page address, the actual value of the i-th accessed page address, and the user portrait to be predicted for the first Two preset steps for training the page prediction model, including:

Using the real value of the i-th visit page address as the i-th visit page calibration value;

Inputting the user portrait to be predicted and the actual value of the i-1th visited page address into the second preset page prediction model to predict the page address of the i-th visited page, and obtain the predicted value of the i-th visited page address;

The second preset page prediction model is trained according to the i-th access page address prediction value and the i-th access page calibration value.
The method for caching page resources according to claim 5, wherein the step of updating the first preset page prediction model according to the trained second preset page prediction model comprises:

Update the model parameters in the first preset page prediction model according to the model parameters in the second preset page prediction model by using a preset model parameter update time, wherein the first preset page prediction model and The model structures of the second preset page prediction models are the same.
A caching device for page resources, wherein the device includes:

The page loading completion signal acquisition module is used to obtain the page loading completion signal of the i-th visit page, and the page loading completion signal carries the user identifier to be predicted and the real value of the i-th visit page address;

A target user portrait acquisition module, configured to acquire a target user portrait according to the user identifier to be predicted in response to the page loading completion signal;

The page address prediction module is configured to use the first preset page prediction model to predict the page address of the i+1th visit according to the real value of the page address of the ith visit and the target user portrait, to obtain the i+1th visit The predicted value of the page address of the second visit;

A cache lookup module, configured to look up the i+1th access page address prediction value in a local cache;

The page resource determination module to be cached is used to obtain the page resource from the server according to the i+1th access page address prediction value to obtain the page resource to be cached when no cached result is found;

A storage module, configured to store the page resource to be cached in the local cache.
A computer device, comprising a memory and a processor, the memory stores a computer program, wherein the processor implements the following method steps when executing the computer program:

Acquiring the page loading completion signal of the i-th visit page, the page loading completion signal carrying the user identifier to be predicted and the real value of the i-th visit page address;

Responding to the page loading completion signal, acquiring a target user portrait according to the user identifier to be predicted;

Using the first preset page prediction model, predicting the page address of the i+1th visit according to the real value of the page address of the ith visit and the target user portrait, to obtain the predicted value of the page address of the i+1th visit;

Searching the predicted value of the i+1th access page address in the local cache;

When the cache result is not found, obtain the page resource from the server according to the i+1th access page address prediction value, and obtain the page resource to be cached;

storing the page resources to be cached in the local cache.
The computer device according to claim 9, wherein the step of obtaining a target user portrait according to the user identifier to be predicted comprises:

calling the page prediction interface, inputting the user identification to be predicted into the user portrait model, and obtaining the target user portrait output by the user portrait model, wherein the user portrait model obtains the user to be profiled according to the user identification to be predicted Data, perform user portrait according to the user data to be portraited, and obtain the target user portrait.
The computer device according to claim 9, wherein the step of searching the i+1th access page address prediction value in a local cache comprises:

Searching the predicted value of the i+1th access page address in the local cache;

When the page address is not found, it is determined that no cached result is found in the local cache;

When the page address is found, obtain the version identifier from the server according to the i+1th access page address prediction value, obtain the version identifier to be cached, and store the i+1th access in the local cache The version identifier of the page resource corresponding to the predicted value of the accessed page address is used as the local cache version identifier;

comparing the version identifier to be cached with the locally cached version identifier;

When the version identifier to be cached is the same as the local cache version identifier, determine that a cached result is found in the local cache;

When the version identifier to be cached is different from the local cache version identifier, it is determined that no cached result is found in the local cache.
The computer device according to claim 9, wherein the step of obtaining page resources from the server according to the i+1th access page address prediction value to obtain page resources to be cached includes:

Generate a page resource acquisition request according to the predicted value of the i+1th access page address, and send the page resource acquisition request to the server;

Acquire the page resource sent by the server according to the page resource acquisition request as the page resource to be cached.
The computer device according to claim 9, wherein, after the step of obtaining the page loading completion signal of the i-th accessed page, further comprising:

Obtain the real value of the page address of the i-1th visited page according to the user identifier to be predicted and the real value of the i-th visited page address, and obtain the real value of the i-1th visited page address;

Obtaining a user portrait to be predicted according to the user identifier to be predicted;

Training a second preset page prediction model according to the actual value of the i-1th visited page address, the true value of the i-th visited page address, and the user portrait to be predicted;

Updating the first preset page prediction model according to the trained second preset page prediction model.
The computer device according to claim 13, wherein, the second preset according to the real value of the i-1th visited page address, the real value of the i-th visited page address and the user portrait to be predicted The steps of training the page prediction model include:

Using the real value of the i-th visit page address as the i-th visit page calibration value;

Inputting the user portrait to be predicted and the actual value of the i-1th visited page address into the second preset page prediction model to predict the page address of the i-th visited page, and obtain the predicted value of the i-th visited page address;

The second preset page prediction model is trained according to the i-th access page address prediction value and the i-th access page calibration value.
A computer-readable storage medium, on which a computer program is stored, wherein, when the computer program is executed by a processor, the following method steps are implemented:

Acquiring the page loading completion signal of the i-th visit page, the page loading completion signal carrying the user identifier to be predicted and the real value of the i-th visit page address;

Responding to the page loading completion signal, acquiring a target user portrait according to the user identifier to be predicted;

Using the first preset page prediction model, predicting the page address of the i+1th visit according to the real value of the page address of the ith visit and the target user portrait, to obtain the predicted value of the page address of the i+1th visit;

Searching the predicted value of the i+1th access page address in the local cache;

When the cache result is not found, obtain the page resource from the server according to the i+1th access page address prediction value, and obtain the page resource to be cached;

storing the page resources to be cached in the local cache.
The computer-readable storage medium according to claim 15, wherein the step of acquiring a target user portrait according to the user identifier to be predicted comprises:

calling the page prediction interface, inputting the user identification to be predicted into the user portrait model, and obtaining the target user portrait output by the user portrait model, wherein the user portrait model obtains the user to be profiled according to the user identification to be predicted Data, perform user portrait according to the user data to be portraited, and obtain the target user portrait.
The computer-readable storage medium according to claim 15, wherein the step of looking up the i+1th access page address prediction value in a local cache comprises:

Searching the predicted value of the i+1th access page address in the local cache;

When the page address is not found, it is determined that no cached result is found in the local cache;

When the page address is found, obtain the version identifier from the server according to the i+1th access page address prediction value, obtain the version identifier to be cached, and store the i+1th access in the local cache The version identifier of the page resource corresponding to the predicted value of the accessed page address is used as the local cache version identifier;

comparing the version identifier to be cached with the locally cached version identifier;

When the version identifier to be cached is the same as the local cache version identifier, determine that a cached result is found in the local cache;

When the version identifier to be cached is different from the local cache version identifier, it is determined that no cached result is found in the local cache.
The computer-readable storage medium according to claim 15, wherein the step of obtaining page resources from the server according to the predicted value of the i+1th access page address, and obtaining the page resources to be cached comprises:

Generate a page resource acquisition request according to the predicted value of the i+1th access page address, and send the page resource acquisition request to the server;

Acquire the page resource sent by the server according to the page resource acquisition request as the page resource to be cached.
The computer-readable storage medium according to claim 15, wherein, after the step of obtaining the page loading completion signal of the i-th accessed page, further comprising:

Obtain the real value of the page address of the i-1th visited page according to the user identifier to be predicted and the real value of the i-th visited page address, and obtain the real value of the i-1th visited page address;

Obtaining a user portrait to be predicted according to the user identifier to be predicted;

Training a second preset page prediction model according to the actual value of the i-1th visited page address, the true value of the i-th visited page address, and the user portrait to be predicted;

Updating the first preset page prediction model according to the trained second preset page prediction model.
The computer-readable storage medium according to claim 19, wherein, according to the actual value of the address of the i-1th accessed page, the actual value of the address of the i-th accessed page, and the user portrait to be predicted for the first Two preset steps for training the page prediction model, including:

Using the real value of the i-th visit page address as the i-th visit page calibration value;

Inputting the user portrait to be predicted and the actual value of the i-1th visited page address into the second preset page prediction model to predict the page address of the i-th visited page, and obtain the predicted value of the i-th visited page address;

The second preset page prediction model is trained according to the i-th access page address prediction value and the i-th access page calibration value.