Application preloading method and device, equipment and storage medium

Abstract

The invention discloses an application preloading method and device, equipment and a storage medium, and relates to the technical field of application management, and the method comprises the steps: obtaining user behavior data of a target user and current network environment data; inputting the user behavior data into a user preference prediction model for preference analysis to obtain user preference information output by the user preference prediction model; inputting the user preference information and the current network environment data into an application recommendation engine for recommendation analysis to obtain an application recommendation list and a resource preloading strategy output by the application recommendation engine; and based on the resource preloading strategy, performing application preloading on the application recommendation list through a preset back-end cache library. According to the application, the application needing to be preloaded is selected through the user preference and the network environment, unnecessary resource loading is avoided, the resource utilization efficiency is improved while the loading time is remarkably shortened, personalized application recommendation is provided for the user, and the personalized requirement of the user is met.

Description

Technical Field

[0001] This application relates to the field of application management technology, and in particular to an application preloading method, apparatus, device and storage medium. Background Technology

[0002] In today's digital age, mini-programs, as a convenient form of mobile application, have been widely used in people's daily lives and work. They can be quickly opened by scanning a code or searching without downloading or installing, providing great convenience to users.

[0003] However, with the widespread use of mini programs, some problems have gradually emerged in their loading mechanism. First, mini programs load slowly, especially when users urgently need to use mini program functions. Long waiting times will reduce user satisfaction with mini programs. Second, resource utilization efficiency is low. Since the loading mechanism of mini programs mainly relies on users' direct access requests, it lacks effective prediction of user behavior and network conditions, resulting in resource waste and seriously affecting user experience. Summary of the Invention

[0004] The main purpose of this application is to provide an application preloading method, apparatus, device and storage medium, which aims to solve the problems of slow loading speed and low resource utilization efficiency of applets in traditional methods.

[0005] To achieve the above objectives, this application proposes an application preloading method, the method comprising:

[0006] Acquire user behavior data and current network environment data of the target user;

[0007] The user behavior data is input into the user preference prediction model for preference analysis to obtain the user preference information output by the user preference prediction model.

[0008] The user preference information and the current network environment data are input into the application recommendation engine for recommendation analysis, and the application recommendation list and resource preloading strategy output by the application recommendation engine are obtained.

[0009] Based on the resource preloading strategy, the application recommendation list is preloaded using a preset backend cache library.

[0010] In one embodiment, the step of inputting the user behavior data into a user preference prediction model for preference analysis to obtain user preference information output by the user preference prediction model includes:

[0011] Feature extraction is performed on the user behavior data to obtain several user behavior features;

[0012] Based on a preset collaborative filtering algorithm and a preset content recommendation algorithm, the user preference prediction model performs matrix decomposition on each of the user behavior features to obtain the user preference information output by the user preference prediction model.

[0013] In one embodiment, the application recommendation engine includes a data integration layer, a model prediction layer, and a resource preloading layer;

[0014] The step of inputting the user preference information and the current network environment data into the application recommendation engine for recommendation analysis, and obtaining the application recommendation list and resource preloading strategy output by the application recommendation engine, includes:

[0015] The data integration layer integrates and analyzes the current network environment data to obtain integrated network environment data.

[0016] The integrated network environment data and the user preference information are input into the model prediction layer for model evaluation, and the model prediction result output by the model prediction layer is obtained.

[0017] The model prediction results are input into the resource preloading layer to obtain the application recommendation list and resource preloading strategy output by the resource preloading layer.

[0018] In one embodiment, the step of preloading applications in the application recommendation list based on the resource preloading strategy using a preset backend cache library includes:

[0019] Obtain user feedback and network changes;

[0020] Based on the user feedback and network changes, update the resource preloading strategy and the application recommendation list;

[0021] Based on the updated resource preloading strategy, the updated application recommendation list is preloaded using the preset backend cache library.

[0022] In one embodiment, updating the resource preloading strategy and the application recommendation list based on the user feedback information and network changes includes:

[0023] The application cache capacity and resource update frequency in the resource preloading strategy are dynamically updated based on the network changes.

[0024] The user feedback information is analyzed to obtain real-time analysis results;

[0025] Based on the real-time analysis results, the application priorities in the application recommendation list are updated.

[0026] In one embodiment, the step of preloading applications based on the updated resource preloading strategy and using a preset backend cache library to preload the updated application recommendation list includes:

[0027] Determine the app ranking in the updated app recommendation list;

[0028] Based on the cache time, cache size, and cache expiration policy in the updated resource preloading strategy, the applications in the updated application recommendation list are loaded into the preset backend cache library according to the application sorting.

[0029] In one embodiment, after preloading the application recommendation list using the preset backend cache library based on the resource preloading strategy, the method further includes:

[0030] Read the application loading request of the target user;

[0031] The application loading request is matched with the preset backend cache library to determine whether the match is successful.

[0032] If so, the application data corresponding to the application loading request is read from the preset backend cache library, and the application data is returned to the target user in response to the user request.

[0033] Furthermore, to achieve the above objectives, this application also proposes an application preloading device, which includes:

[0034] The data acquisition module is used to acquire user behavior data of the target user and current network environment data;

[0035] The preference prediction module is used to input the user behavior data into the user preference prediction model for preference analysis, and obtain the user preference information output by the user preference prediction model.

[0036] The application recommendation module is used to input the user preference information and the current network environment data into the application recommendation engine for recommendation analysis, and obtain the application recommendation list and resource preloading strategy output by the application recommendation engine.

[0037] The application loading module is used to preload the application recommendation list based on the resource preloading strategy and through a preset backend cache library.

[0038] In addition, to achieve the above objectives, this application also proposes an application preloading device, the device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the application preloading method as described above.

[0039] In addition, to achieve the above objectives, this application also proposes a storage medium, which is a computer-readable storage medium, on which a computer program is stored, and which, when executed by a processor, implements the steps of the application preloading method described above.

[0040] In addition, to achieve the above objectives, this application also provides a computer program product, which includes a computer program that, when executed by a processor, implements the steps of the application preloading method described above.

[0041] This application provides an application preloading method, apparatus, device, and storage medium. The application preloading method acquires user behavior data and current network environment data of a target user, then inputs the user behavior data into a user preference prediction model for preference analysis, obtaining user preference information output by the model. This user preference information and the current network environment data are then input into an application recommendation engine for recommendation analysis, resulting in an application recommendation list and resource preloading strategy output by the engine. Based on the resource preloading strategy, the application recommendation list is preloaded using a pre-set backend cache library. This preloading strategy preloads applications that the user may access into the cache. When a user makes a request, the application can be quickly loaded directly from the cache, significantly reducing loading time. Simultaneously, it intelligently selects applications to be preloaded based on user preferences and the network environment, avoiding unnecessary resource loading, saving network bandwidth, improving resource utilization efficiency, and providing personalized application recommendations to meet the user's individual needs. Furthermore, since the applications are preloaded into the cache, user requests can be directly responded to by the cache, reducing server load and response time. Attached Figure Description

[0042] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0043] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0044] Figure 1 A flowchart illustrating the application of the preloading method in Embodiment 1 of this application;

[0045] Figure 2A flowchart illustrating the data acquisition process for the application of the preloading method in this application;

[0046] Figure 3 A flowchart illustrating the user behavior analysis process provided for the application of the preloading method in this application;

[0047] Figure 4 A flowchart illustrating the preloading strategy generation process for applying the preloading method in this application;

[0048] Figure 5 A flowchart illustrating dynamic cache management provided for the application of the preloading method in this application;

[0049] Figure 6 A simplified flowchart illustrating the preloading method used in this application;

[0050] Figure 7 A flowchart illustrating the application caching process provided for the application preloading method in this application;

[0051] Figure 8 System architecture diagram provided for the application of the preloading method in this application;

[0052] Figure 9 This is a schematic diagram of the module structure of the preloading device used in an embodiment of this application;

[0053] Figure 10 This is a schematic diagram of the device structure of the hardware operating environment involved in the application of the preloading method in the embodiments of this application.

[0054] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0055] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application.

[0056] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0057] It should be noted that the executing entity in this embodiment can be a computing service device with data processing, network communication, and program execution functions, such as a tablet computer, personal computer, or mobile phone, or an electronic device, big data service platform, or application preloading system capable of performing the above functions. The following description uses an application preloading system as an example to illustrate this embodiment and the subsequent embodiments.

[0058] Based on this, embodiments of this application provide an application preloading method, referring to... Figure 1 , Figure 1This is a flowchart illustrating the application of the preloading method in Embodiment 1 of this application.

[0059] In this embodiment, the application preloading method includes steps S11 to S14:

[0060] Step S11: Obtain user behavior data and current network environment data of the target user;

[0061] It should be noted that the target user refers to the terminal device user who needs to perform application preloading service; the user behavior data refers to the quantifiable records generated during the interaction between the user and the mini program, including browsing history (such as the sequence of mini program IDs visited), search history (keywords and frequency), click behavior (button trigger location and timestamp), dwell time (duration of a single session), and usage frequency (number of visits per unit time), etc.

[0062] It should be further noted that the current network environment data refers to the real-time monitored network status parameters of user equipment, including bandwidth (data transmission volume per unit time, in Mbps), latency (round-trip time of data packets, in ms), and packet loss rate (percentage of data packets that failed to transmit).

[0063] Specifically, the system collects user behavior logs through a front-end event tracking SDK and dynamically captures network metrics by calling network diagnostic interfaces (such as the WebSocket speed test protocol). For example, bandwidth is calculated by downloading a fixed-size test packet to determine the transmission rate, and latency is measured by the response time difference of HTTP requests. This multi-dimensional data fusion provides input support for subsequent intelligent recommendations, avoiding prediction biases caused by single data sets in traditional technologies.

[0064] For example, when a user uses an e-commerce mini-program, the system records the duration of their browsing of the product details page (stay time = 120 seconds), the frequency of searching for the keyword "smartphone" (3 times / day), and also detects the current 4G network bandwidth (50Mbps) and latency (80ms), and encapsulates it into structured data packets for transmission to the backend analysis module.

[0065] In one embodiment, the category of a user's mini-program can be extracted based on the type of mini-program accessed by the user, such as shopping, social, news, etc. Metadata of each mini-program, including category information, can be obtained through an API interface. Usage frequency can be measured by counting the number of times a user accesses each mini-program, such as recording the timestamp of each access and counting the number of accesses within a certain time range (e.g., one day, one week, or one month). Dwell time can be measured by recording the length of time a user spends in each mini-program, such as recording the start time when the mini-program loads and the end time when the mini-program closes, and calculating the difference between the two. Total dwell time can be measured by recording the total dwell time in each mini-program, such as summing up the dwell time of each access to obtain the total dwell time. Capturing bandwidth can be estimated by uploading or downloading fixed-size data packets. For example, WebSocket or other HTTP requests can be used to upload or download data, and then the bandwidth can be calculated based on the required time. Capturing latency can be achieved by measuring the round-trip time (RTT) from the client to the server, such as by sending an HTTP request to the server and recording the time difference between the request being sent and the response arriving. Capturing packet loss rate can be estimated by sending a series of data packets and counting the number of packets received, such as by sending data packets using WebSocket or UDP protocols and counting the number of packets received. Evaluating the quality of network conditions can be done based on bandwidth, latency, and packet loss rate. For example, a set of thresholds can be defined to evaluate network conditions. See reference [link / reference]. Figure 2 .

[0066] It should be noted that the above-mentioned data collection, data security, data integration, and data privacy are to ensure that relevant laws and regulations are followed when collecting data, that the user's explicit consent is obtained, that the security of user data is protected, that data is encrypted during transmission, that user privacy is respected, that unnecessary personal information is not collected, and that data is anonymized, so as to compliantly and effectively extract relevant behavioral characteristics of users in the mini-program from mobile phone data.

[0067] Step S12: Input the user behavior data into the user preference prediction model for preference analysis to obtain the user preference information output by the user preference prediction model;

[0068] It should be noted that the user preference prediction model refers to an algorithmic model capable of predicting user interests and preferences based on user behavior data, which can be implemented through techniques such as machine learning and data mining. The user preference information refers to information obtained after model analysis regarding application categories and functions that the user may be interested in.

[0069] Specifically, feature extraction is performed on the user behavior data to obtain several user behavior features. Then, based on a preset collaborative filtering algorithm and a preset content recommendation algorithm, matrix decomposition is performed on each of the user behavior features through the user preference prediction model to obtain the user preference information output by the user preference prediction model.

[0070] Step S13: Input the user preference information and the current network environment data into the application recommendation engine for recommendation analysis, and obtain the application recommendation list and resource preloading strategy output by the application recommendation engine;

[0071] It should be noted that the application recommendation engine refers to a system component that generates an application recommendation list and resource preloading strategy based on user preference information and network environment data. The resource preloading strategy refers to an application resource loading plan formulated based on user preferences and network conditions, including parameters such as application cache capacity and resource update frequency. The application recommendation list refers to a list of applications recommended to the user, where each application has a priority.

[0072] Specifically, the data integration layer integrates and analyzes the current network environment data to obtain integrated network environment data. Then, the integrated network environment data and the user preference information are input into the model prediction layer for model evaluation to obtain the model prediction result output by the model prediction layer. The model prediction result is then input into the resource preloading layer to obtain the application recommendation list and resource preloading strategy output by the resource preloading layer.

[0073] Step S14: Based on the resource preloading strategy, preload the application recommendation list using a preset backend cache library.

[0074] It should be noted that the backend cache library refers to a system component used to store preloaded resources. Application preloading refers to the process of loading application resources into the cache before a user actually requests the application. This is achieved through the backend cache library, enabling faster service delivery and reducing waiting time when a user requests the application.

[0075] Specifically, user feedback information and network changes are obtained, and then the resource preloading strategy and the application recommendation list are updated based on the user feedback information and the network changes. Based on the updated resource preloading strategy, the updated application recommendation list is preloaded through a preset backend cache library.

[0076] This embodiment acquires user behavior data and current network environment data of the target user, then inputs the user behavior data into a user preference prediction model for preference analysis, obtaining user preference information output by the model. This user preference information, along with the current network environment data, is then input into an application recommendation engine for recommendation analysis, resulting in an application recommendation list and resource preloading strategy. Based on the resource preloading strategy, applications on the recommendation list are preloaded using a pre-defined backend cache library. This preloading strategy preloads applications that the user may access into the cache. When a user makes a request, the application can be quickly loaded directly from the cache, significantly reducing loading time. Simultaneously, it intelligently selects applications to be preloaded based on user preferences and the network environment, avoiding unnecessary resource loading, saving network bandwidth, improving resource utilization efficiency, and providing personalized application recommendations to meet the user's individual needs. Furthermore, since the applications are preloaded into the cache, user requests can be directly responded to by the cache, reducing server load and response time.

[0077] In one feasible implementation, the step of inputting the user behavior data into a user preference prediction model for preference analysis to obtain user preference information output by the user preference prediction model includes:

[0078] Step S21: Extract features from the user behavior data to obtain several user behavior features;

[0079] It should be noted that the feature extraction process refers to identifying and extracting key information that can represent user behavior patterns from user behavior data. These user behavior features refer to attributes that reflect specific user habits or preferences, such as user click frequency, access time periods, preferred mini-program categories, and dwell time.

[0080] Specifically, feature extraction is performed on the user behavior data, such as the category of mini-programs browsed by the user (using one-hot encoding to convert the category into numerical features), frequency, and time period (the time period during which the user accesses the mini-program (morning, afternoon, evening, etc.)). This yields several user behavior features. The feature extraction process can include sub-steps such as data cleaning, feature selection, and feature construction. Data cleaning removes noise and incomplete data; feature selection chooses the most representative features from a large amount of raw data; and feature construction creates new features or transforms existing features to better capture user behavior. In one possible implementation, feature extraction can be achieved through data mining techniques, including but not limited to statistical analysis and machine learning methods, to identify important patterns and trends in user behavior.

[0081] Step S22: Based on the preset collaborative filtering algorithm and the preset content recommendation algorithm, the user preference prediction model is used to perform matrix decomposition on each of the user behavior features to obtain the user preference information output by the user preference prediction model.

[0082] It should be noted that the collaborative filtering algorithm refers to making recommendations by analyzing the similarity between users or items, and it can be divided into user-based collaborative filtering and item-based collaborative filtering. The content recommendation algorithm, on the other hand, makes recommendations based on users' past behavior and preferred content, typically involving the analysis of content features and modeling of user preferences. Matrix factorization is a mathematical technique used in recommendation systems to discover potential relationships between users and items. By decomposing the user-item rating matrix into two lower-dimensional matrices, it reveals the latent characteristics of users and items.

[0083] Specifically, based on a preset collaborative filtering algorithm (such as user-user collaborative filtering: recommending content based on the similarity between users; or item-item collaborative filtering: recommending content based on the similarity between items) and a preset content recommendation algorithm (using text analysis or image analysis techniques to extract content features, and then recommending similar content based on these features), the user preference prediction model performs matrix decomposition on each of the user behavior features to obtain the user preference information output by the user preference prediction model. The matrix decomposition process may include selecting an appropriate decomposition method (such as Singular Value Decomposition (SVD), Non-negative Matrix Factorization (NMF), etc., to discover potential relationships between users and items), determining the dimensions of the decomposition, and performing the decomposition calculation, etc., which can be referred to... Figure 3 .

[0084] Additionally, the training process of the user preference prediction model includes: collecting data such as the user's browsing history, click behavior, search records, and dwell time; extracting useful features from the collected data, such as browsing time, browsing frequency, and content category; and selecting appropriate machine learning algorithms based on the characteristics of the problem. For example, Convolutional Neural Networks (CNNs) are suitable for image-based content analysis, Recurrent Neural Networks (RNNs) or Long Short-Term Memory Networks (LSTMs) are suitable for time-series data analysis, and Transformer models are suitable for text-based data analysis. These can be set according to the actual situation. The model is then constructed using methods such as logistic regression, support vector machines, or random forests. Logistic regression is simple, highly interpretable, and suitable for binary classification problems, such as predicting whether a user will visit a certain app. Support vector machines (SVMs) are suitable for small sample data and have good generalization ability, such as classification problems, especially when the data has non-linear boundaries. Random forests can handle high-dimensional data, are less prone to overfitting, and are suitable for classification and regression problems, especially when there is correlation between features.

[0085] Furthermore, the model is trained using the training dataset and evaluated using the test dataset. The model is evaluated using parameters such as accuracy (the consistency between the predicted result and the actual result), recall (the probability of correctly predicting a positive example), precision (the proportion of positive examples that are actually positive examples), F1 score (the harmonic mean of precision and recall), ROC curve, and AUC value (a common method for evaluating classifier performance). Finally, the trained model is used to predict users' interests and provide personalized recommendations accordingly.

[0086] This embodiment extracts features from the user behavior data to obtain several user behavior features. Then, based on a preset collaborative filtering algorithm and a preset content recommendation algorithm, the user preference prediction model performs matrix decomposition on each of the user behavior features to obtain the user preference information output by the user preference prediction model. Thus, feature extraction and matrix decomposition more accurately capture user preferences, improve the accuracy of application recommendations, and provide users with application recommendations that are more tailored to their individual needs, enhancing the user experience. This also allows for more efficient allocation of resources, such as bandwidth and storage space, to support applications that users may be interested in. At the same time, user preference information can be dynamically updated as user behavior changes, enabling the recommendation system to adapt to changes in user preferences and maintain the timeliness of recommendations.

[0087] In one feasible implementation, the application recommendation engine includes a data integration layer, a model prediction layer, and a resource preloading layer; the step of inputting the user preference information and the current network environment data into the application recommendation engine for recommendation analysis, and obtaining the application recommendation list and resource preloading strategy output by the application recommendation engine, includes:

[0088] Step S31: The current network environment data is integrated and analyzed through the data integration layer to obtain integrated network environment data;

[0089] It should be noted that the data integration layer refers to the system module responsible for cleaning, classifying, and aggregating the raw network environment data; the integrated network environment data refers to the structured network status description after standardization, such as: network quality level (excellent / good / poor, based on a comprehensive score of bandwidth, latency, and packet loss rate); available bandwidth segments (e.g., 0-10Mbps is "low bandwidth", 10-50Mbps is "medium bandwidth"); real-time load status (e.g., the percentage of users currently connected to the base station), which is not limited here and can be set according to the actual situation.

[0090] Specifically, the data integration layer integrates and analyzes the current network environment data to obtain integrated network environment data. The data integration layer performs the following operations: data cleaning: removing outliers (e.g., invalid probe data with latency > 1000ms); feature normalization: mapping parameters of different dimensions (e.g., bandwidth Mbps, latency ms) to the [0,1] interval; dynamic grading: classifying network states according to preset thresholds (e.g., marking bandwidth < 5Mbps as "weak network mode") to address the heterogeneity problem of the original network data. For example, the differences between 4G / 5G / Wi-Fi indicators can be uniformly converted into a universal quality score for easier subsequent model processing. (See reference...) Figure 4 .

[0091] In addition, in one possible implementation, the data integration layer uses a sliding window algorithm (window size = 5 minutes) to smooth out instantaneous fluctuations and avoid misjudgments caused by network jitter.

[0092] Step S32: Input the integrated network environment data and the user preference information into the model prediction layer for model evaluation, and obtain the model prediction result output by the model prediction layer;

[0093] It should be noted that the model prediction layer refers to the decision-making module based on machine learning algorithms. Its input is the integrated network data and user preference vector, and its output is the model prediction result. The model prediction result includes: predicted access probability (e.g., access probability of mini-program A = 85%); resource requirement assessment (e.g., mini-program B needs to preload JS / CSS resource package size = 3MB); network adaptability tags (e.g., "suitable for loading on weak networks"), etc.

[0094] Specifically, the integrated network environment data and the user preference information are input into the model prediction layer for model evaluation. The model prediction layer employs the following techniques: multi-feature fusion: the user preference information (e.g., shopping category weight 0.9) is concatenated with the integrated network environment data (e.g., bandwidth level "low") to form a joint feature vector; probabilistic prediction: gradient boosting tree (GBDT) is used to calculate the access probability of each mini-program; resource constraint filtering: mini-programs that cannot be loaded under the current network conditions are removed (e.g., filtering high-definition video applications when bandwidth < 5Mbps), resulting in the model prediction result output by the model prediction layer.

[0095] Furthermore, by introducing network adaptive constraints, the resource waste caused by traditional recommendation systems relying solely on user preferences can be avoided. In one possible implementation, the model prediction layer employs a dynamic weight adjustment mechanism: when the network quality is poor, the weight of the network state features is increased from the default 30% to 60%, which can be set according to the actual situation and is not limited here.

[0096] Step S33: Input the model prediction results into the resource preloading layer to obtain the application recommendation list and resource preloading strategy output by the resource preloading layer.

[0097] It should be noted that the resource preloading layer is a system component that executes the final decision, generating actionable caching instructions based on model prediction results. The application recommendation list refers to a list of mini-program IDs arranged in descending order of access probability (e.g., [A, B, C]); the resource preloading strategy includes cache capacity allocation (e.g., reserving 10MB of memory for mini-program A), update frequency (e.g., checking the version every 15 minutes), expiration rules (e.g., LRU eviction policy), etc.

[0098] Specifically, the model prediction results are input into the resource preloading layer to obtain the application recommendation list and resource preloading strategy output by the resource preloading layer. In one embodiment, the resource preloading layer executes the following logic: priority sorting: applications with a prediction probability > 80% are marked as "high priority"; dynamic quota allocation: resources are allocated proportionally based on the current remaining cache capacity (e.g., Redis remaining memory = 100MB); network adaptation optimization: resource compression is enabled under weak network conditions (e.g., converting images to WebP format). In addition, the resource preloading layer can dynamically adjust the preloading order based on historical hit rates (e.g., the cache hit rate of mini-program A = 90%).

[0099] This embodiment integrates and analyzes the current network environment data through the data integration layer to obtain integrated network environment data. This integrated network environment data, along with user preference information, is then input into the model prediction layer for model evaluation, yielding the model prediction results. These results are then input into the resource preloading layer to obtain an application recommendation list and resource preloading strategy. By integrating network environment data, the system can identify the user's current network status (e.g., weak 4G signal / high-speed Wi-Fi), avoiding the recommendation of incompatible applications (e.g., filtering high-definition video applications in weak network conditions), and reducing unnecessary recommendations. The system integrates user preferences to significantly improve preloading hit rate. The data integration layer standardizes raw network data into quantifiable metrics (such as "network quality level"), eliminating interference from different network types (5G / Wi-Fi) and providing a unified input for the model. The model prediction layer uses machine learning to quantify prediction probabilities, prioritizing the loading of high-probability access mini-programs to avoid cache waste caused by the "broad-based" approach in traditional solutions. The resource preloading layer dynamically adjusts strategies based on network conditions (such as compressing resources during weak network conditions) to achieve real-time balance between cache capacity and network conditions. By preloading high-probability applications, users can directly read resources from the cache after clicking, reducing user waiting time and lowering server load.

[0100] In one feasible implementation, the step of preloading applications from the application recommendation list based on the resource preloading strategy using a preset backend cache library includes:

[0101] Step S41: Obtain user feedback information and network changes;

[0102] It should be noted that the user feedback information refers to feedback data generated by users during application use, such as click-through rates, dwell time, ratings, and reviews, reflecting user satisfaction and interest in the application. The network changes refer to the dynamic changes in the network environment of the user's device, including changes in bandwidth, latency, and packet loss rate. These changes affect application loading speed and user experience, allowing for real-time adjustments to resource preloading strategies and application recommendation lists to better meet user needs and optimize resource utilization efficiency.

[0103] Specifically, the user feedback information can be collected through a user behavior monitoring module, for example, by recording user actions within the application (such as clicks, pauses, etc.) to obtain data such as click-through rate and dwell time. Network changes can be monitored in real time through a network status monitoring module, for example, by periodically sending test data packets to measure metrics such as bandwidth, latency, and packet loss rate. In one possible implementation, the system will set a monitoring cycle, for example, monitoring the network status every minute or every 5 minutes, and collecting user feedback information in real time.

[0104] For example, suppose a user spends a long time using a certain mini-program and has a high click-through rate, indicating that the user is interested in that mini-program. Simultaneously, the system detects that the user's current network bandwidth has decreased from a high level to a low level.

[0105] Step S42: Based on the user feedback information and the network changes, update the resource preloading strategy and the application recommendation list;

[0106] It should be noted that the purpose of updating the resource preloading strategy and application recommendation list is to dynamically adjust the preloaded content and strategy based on user feedback and network changes, so as to ensure that the preloaded applications can better meet user needs and can be loaded efficiently in the current network environment.

[0107] Specifically, the application cache capacity and resource update frequency in the resource preloading strategy are dynamically updated based on network changes. This is followed by analysis of user feedback information to obtain real-time analysis results, which are then used to update the application priorities in the application recommendation list. For example, suppose user feedback shows that a user has a high click-through rate and dwell time on a certain news mini-program, indicating high user interest. Simultaneously, network changes show that the user's current network bandwidth is low. Based on this information, the system will reduce the cache capacity and decrease the resource update frequency to adapt to the low-bandwidth environment, and increase the priority of the news mini-program to ensure it is preloaded into the cache first. (See reference...) Figure 5 .

[0108] Step S43: Based on the updated resource preloading strategy, preload the updated application recommendation list using the preset backend cache library.

[0109] It should be noted that the updated resource preloading strategy refers to the application resource loading plan adjusted based on user feedback and network changes, while the updated application recommendation list refers to the application list adjusted based on user interests. The purpose of preloading the updated application recommendation list using a pre-defined backend cache library is to ensure that the preloaded applications better meet user needs and load quickly under the current network conditions, thereby improving the user experience.

[0110] Additionally, it's important to note that the backend cache library is a system component used to store preloaded resources. It can dynamically adjust the content in the cache based on the resource preloading strategy. In one possible implementation, the system loads applications from the updated application recommendation list into the backend cache library in priority order, based on the cache time, cache size, and cache expiration policy in the updated resource preloading strategy. For example, for higher-priority applications, the system will allocate more cache space and ensure a longer cache time.

[0111] For example, in one specific implementation, suppose a shopping mini-program has been given higher priority in the updated application recommendation list, and the updated resource preloading strategy specifies a larger cache capacity and a shorter cache expiration time. Based on these strategies, the system will prioritize loading the resources of this shopping mini-program into the backend cache library, ensuring fast loading when the user requests the application, thereby improving the user experience.

[0112] Specifically, the order of applications in the updated application recommendation list is determined, and then, based on the cache time, cache size, and cache expiration policy in the updated resource preloading strategy, the applications in the updated application recommendation list are loaded into the preset backend cache library according to the application order. (See reference...) Figure 6 .

[0113] This embodiment acquires user feedback and network changes, and then updates the resource preloading strategy and the application recommendation list based on the user feedback and network changes. Based on the updated resource preloading strategy, applications on the updated application recommendation list are preloaded through a preset backend cache library. Furthermore, by analyzing user feedback (such as skipped recommendations, click-through rate, and dwell time), the priority of the recommendation list is dynamically adjusted, and the preloading strategy is optimized in real time. This accurately responds to user behavior and adapts to network fluctuations, effectively improving resource hit rate, ensuring a smooth experience, and maximizing resource utilization efficiency by dynamically adapting to network changes.

[0114] In one feasible implementation, updating the resource preloading strategy and the application recommendation list based on the user feedback information and network changes includes:

[0115] Step S51: Dynamically update the application cache capacity and resource update frequency in the resource preloading strategy based on the network changes.

[0116] It should be noted that the application cache capacity refers to the storage space allocated by the system for preloading application resources, while the resource update frequency refers to the frequency at which the application resources in the cache are updated to adapt to changes in the network environment, ensure the rationality and effectiveness of the preloading strategy, and optimize resource utilization efficiency.

[0117] Specifically, the application cache capacity and resource update frequency in the resource preloading strategy are dynamically updated based on network changes. In one possible implementation, when network bandwidth is low, the system reduces the application cache capacity to avoid consuming excessive network bandwidth, thereby ensuring more efficient use of network resources. Simultaneously, the system reduces the resource update frequency to minimize unnecessary data transmission, adapting to low-bandwidth environments. Conversely, when network bandwidth is high and stable, the system can appropriately increase the application cache capacity and improve the resource update frequency to ensure users receive the latest version of application resources. For example, when a user switches from a Wi-Fi environment to a 4G network, the system dynamically adjusts the cache capacity and update frequency according to changes in network bandwidth, ensuring the flexibility and adaptability of the preloading strategy.

[0118] Step S52: Analyze the user feedback information to obtain real-time analysis results;

[0119] It should be noted that the real-time analysis results refer to the results obtained by instantly analyzing user feedback information, which are used to guide subsequent application recommendations and resource preloading strategy adjustments.

[0120] Specifically, in one possible implementation, the system monitors and analyzes user data such as click-through rate and dwell time in real time. For example, if an application has a high click-through rate and a long user dwell time, it indicates that the user has a high interest in that application. The system quantifies this level of interest into real-time analysis results, such as by calculating an interest score for each application. These real-time analysis results are used to adjust the subsequent application recommendation list to ensure that the recommended applications better meet the user's immediate needs. For example, the system may prioritize recommending applications with higher user interest scores to the user based on the real-time analysis results.

[0121] Step S53: Based on the real-time analysis results, update the application priorities in the application recommendation list.

[0122] It should be noted that the application priority reflects the sorting order of the applications in the recommendation list. The higher the priority, the higher the position of the application in the recommendation list, so as to ensure that the recommended applications better match the user's immediate interests and needs, thereby improving the accuracy of recommendations and user experience.

[0123] Specifically, the system adjusts the priority of apps in the recommendation list based on real-time analysis results. For example, if real-time analysis shows that a user has a high interest score in a news app, the system will increase its priority, placing it higher in the recommendation list. Conversely, if user feedback indicates a low user interest in an app, the system will decrease its priority. This dynamic adjustment mechanism ensures that recommended apps always align with user interests, thereby improving the effectiveness of the recommendation system and user satisfaction. For instance, if a user frequently clicks on a shopping app and stays on it for a considerable period, the system will increase its priority based on this feedback, ensuring it is shown to the user more frequently in subsequent recommendations.

[0124] This embodiment dynamically updates the application cache capacity and resource update frequency in the resource preloading strategy based on network changes. It then analyzes user feedback to obtain real-time analysis results. Based on these results, it updates the application priorities in the application recommendation list. By dynamically adjusting the application cache capacity and resource update frequency, the system can rationally allocate resources according to current network conditions (such as bandwidth and latency). When network conditions are poor, the cache capacity and update frequency are reduced to avoid lag or loading failures caused by excessive resource loading. This ensures that users can quickly obtain the content they need when requesting applications, significantly reducing waiting time and enabling rapid response to user needs. The system can also adjust the application priorities in the application recommendation list in a timely manner, prioritizing applications that users are more interested in. Since traditional preloading strategies are usually based on fixed rules, they may load a large amount of unnecessary resources when users are not interested or network conditions are not suitable, leading to resource waste. Therefore, by dynamically updating the strategy, the system can flexibly adjust according to actual needs and network conditions, avoiding invalid resource loading and improving resource utilization efficiency. Simultaneously, it dynamically adapts to network changes, ensuring that the preloading strategy is always in an optimal state, avoiding resource loading failures or delays caused by network changes.

[0125] In one feasible implementation, the step of preloading applications from the updated application recommendation list using the preset backend cache library based on the updated resource preloading strategy includes:

[0126] Step S61: Determine the application ranking in the updated application recommendation list;

[0127] It should be noted that the updated app recommendation list refers to the list of apps adjusted based on user feedback and network changes, in which the priority of each app has been reordered according to user interests and network conditions.

[0128] It should be further explained that the application sorting refers to arranging the application recommendation list according to the application priority, with higher priority applications listed first and lower priority applications listed last, to ensure that the applications that users are most likely to access are loaded into the cache first during the preloading process, thereby improving the user experience.

[0129] Specifically, the system sorts the recommended apps based on real-time analysis results. For example, if an app has a high user interest score (e.g., high click-through rate, long dwell time), its priority is increased, placing it higher on the recommendation list. Conversely, if an app has a low user interest score, its priority is decreased, placing it lower on the recommendation list. Furthermore, the system can further optimize app ranking by combining app type (e.g., news, shopping, social media) and user habits (e.g., whether users tend to use certain types of apps during certain time periods). For example, if users use news apps more frequently at night, the system will appropriately increase the priority of news apps at night, and so on, without limitation.

[0130] Step S62: Based on the cache time, cache size, and cache expiration policy in the updated resource preloading strategy, load the applications in the updated application recommendation list into the preset backend cache library according to the application sorting.

[0131] It should be noted that the updated resource preloading strategy refers to the preloading strategy adjusted based on network changes and user feedback, which includes parameters such as cache duration, cache size, and cache expiration policy. Specifically, cache duration refers to the length of time application resources are retained in the cache; cache size refers to the amount of storage space allocated to the cache; and the cache expiration policy refers to the rules that determine when resources in the cache are updated or replaced.

[0132] Specifically, the system loads applications from the application recommendation list into the backend cache library according to the parameters in the updated resource preloading strategy, sorted by application order. In one embodiment, anchor points are set on the mini-program page. When a user triggers an anchor point, the mini-program passes the anchor point and user data to the backend interface via an API call. The backend interface binds the anchor point and user relationship and stores it in the Redis database (the default backend cache library). (See reference...) Figure 7 For example, for high-priority applications, the system allocates more cache space (cache size) and sets shorter cache expiration times to ensure users can quickly access the latest version of application resources. For low-priority applications, the system may allocate less cache space and set longer cache expiration times to conserve cache resources. Furthermore, the system periodically checks cached resources for expiration based on cache expiration parameters and promptly updates expired resources to ensure they are always valid. For instance, if an application's cache expiration time is set to 24 hours, the system will automatically update the application's cached resources after 24 hours to ensure users receive the latest content upon request.

[0133] Furthermore, front-end caching components work by storing frequently accessed data on the front-end device (such as a browser) for quick retrieval in subsequent requests, reducing server requests and thus improving page loading speed and performance. Redis caching, on the other hand, stores data and user relationships in server-side memory. When there are many users, multiple Redis caches can be deployed to address this issue, which front-end caching components cannot solve. Moreover, data cached by front-end components cannot be shared between them, whereas data and user relationships cached by the Redis server can be shared between different users. This sharing occurs when users call a backend API to retrieve the cached data.

[0134] This embodiment determines the application ranking in the updated application recommendation list, and then loads the applications in the updated application recommendation list into the preset backend cache library according to the application ranking based on the cache time, cache size, and cache expiration policy in the updated resource preloading strategy. By preloading the applications that users are most likely to access into the cache, the system can significantly reduce the waiting time for users when requesting applications, achieve fast response, and improve the overall user experience. At the same time, the application ranking is dynamically adjusted based on user interests and real-time feedback, ensuring that users can access the applications they are most interested in more quickly, thereby providing more personalized services. It also ensures that the resources in the cache are up-to-date and meet the user's current needs, avoiding the storage of a large number of outdated or useless resources in the cache, thereby optimizing the cache utilization efficiency, avoiding unnecessary resource loading, saving network bandwidth and storage space, and improving resource utilization efficiency.

[0135] In one feasible implementation, after preloading the application recommendation list based on the resource preloading strategy using a preset backend cache library, the method further includes:

[0136] Step S71: Read the application loading request of the target user;

[0137] It should be noted that the application loading request refers to the loading instruction for a certain application initiated by the user through the operation interface (such as clicking an icon, entering keywords, etc.).

[0138] Specifically, in one possible implementation, the system receives the user's application loading request through the user interface (such as the mini-program's launch page, application list, etc.). For example, when a user clicks the icon of a news mini-program, the system captures this action and parses it into a specific application loading request. The system records the request time, the requested application identifier (such as the mini-program's ID), and the user's device information for subsequent accurate matching and processing.

[0139] Step S72: Match the application loading request with the preset backend cache library and determine whether the match is successful;

[0140] It should be noted that the preset backend cache library is a system component that stores preloaded application data, which includes application resources that users may access.

[0141] Specifically, the application loading request is matched with the preset backend cache library to determine whether the match is successful. A successful match means that the application data requested by the user exists in the cache library and can be read directly from the cache; if the match fails, the data needs to be obtained from other data sources (such as servers).

[0142] Furthermore, the system searches a pre-defined backend cache based on the application identifier (such as the mini-program's ID) in the application loading request. For example, if a user requests a news mini-program, the system will check if the latest data for that news mini-program is stored in the cache. The matching operation can be implemented using efficient data structures such as hash tables and indexed databases to ensure that the matching process is completed quickly. If the match is successful, the system will record the matching result and proceed to the next step; if the match fails, the system will record the reason for the failure and initiate the process of loading data from the server.

[0143] Step S73: If yes, then read the application data corresponding to the application loading request from the preset backend cache library, and return the application data to the target user to respond to the user request.

[0144] It should be noted that reading application data from the preset backend cache library means that after a successful match, the system retrieves the application resources requested by the user from the cache library to quickly respond to the user's request, reduce the user's waiting time, and improve the user experience.

[0145] Specifically, upon successful matching, the system immediately retrieves the corresponding application data from a pre-set backend cache library. For example, if the data for a news-related mini-program requested by the user is already stored in the cache library, the system will extract the data from the cache library and transmit it to the user's device via the network. During transmission, the system ensures data integrity and security, such as protecting user data through encrypted transmission. After receiving the data, the user's device renders and displays it to the user, thus completing the entire application loading process. This process is typically much faster than loading data directly from the server because the cached data is pre-loaded and stored locally or on a server close to the user, reducing network latency and data transmission time.

[0146] This embodiment reads the application loading request from the target user and matches it with a preset backend cache library. If a match is found, the system reads the application data corresponding to the loading request from the preset backend cache library and returns the data to the target user to respond to the user's request. By directly reading application data from the preset backend cache library, the system can quickly respond to user loading requests, significantly reducing user waiting time. Since users do not need to wait for the application to reload from the server, they experience a smoother user experience. Furthermore, the cached data is pre-loaded and optimized, reducing stuttering and loading failures caused by network latency or instability, improving application stability and reliability. This avoids repeatedly loading the same data from the server, reducing unnecessary network traffic consumption. This saves users on data charges and helps alleviate network congestion. Even when the user's network is unstable or bandwidth is low, the system can still quickly provide application data through caching, ensuring normal application use and improving system robustness. Even if the server experiences a brief failure or maintenance, users can still access frequently used application data through caching, reducing the possibility of service interruption.

[0147] For example, to help understand the implementation process of the application preloading method, please refer to... Figure 8 , Figure 8 The system architecture diagram provided for the application of the preloading method in this application.

[0148] Specifically, the diagram illustrates the system module architecture of a mini-program loading optimization system based on intelligent recommendation. This system consists of four main modules: a user behavior analysis module, a network status monitoring module, an intelligent recommendation engine module, and a dynamic cache management module. The user behavior analysis module collects user data, extracts feature values, and analyzes user interests to predict mini-programs the user might be interested in. The network status monitoring module monitors the user's network environment in real time, including metrics such as bandwidth and latency, and transmits this data to the intelligent recommendation engine. The intelligent recommendation engine module combines user behavior analysis results and network status monitoring data to build a predictive model and preload resources to preload mini-program resources that the user might access. Finally, the dynamic cache management module adjusts the caching strategy based on user feedback and network changes, updating the cache strategy and resources to ensure efficient resource utilization. The entire system aims to improve mini-program loading speed and enhance user experience while conserving resources through intelligent prediction and dynamic management.

[0149] It should be noted that the examples in the figure are only for understanding this application and do not constitute a limitation on the application of the preloading method. Any simple modifications based on this technical concept are within the protection scope of this application.

[0150] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0151] This application also provides an application preloading device, please refer to... Figure 9 The application preloading device includes:

[0152] Data acquisition module 91 is used to acquire user behavior data of the target user and current network environment data;

[0153] The preference prediction module 92 is used to input the user behavior data into the user preference prediction model for preference analysis, and obtain the user preference information output by the user preference prediction model.

[0154] Application recommendation module 93 is used to input the user preference information and the current network environment data into the application recommendation engine for recommendation analysis, and obtain the application recommendation list and resource preloading strategy output by the application recommendation engine;

[0155] Application loading module 94 is used to preload the application recommendation list based on the resource preloading strategy and through a preset backend cache library.

[0156] The application preloading device is also used for:

[0157] Feature extraction is performed on the user behavior data to obtain several user behavior features;

[0158] Based on a preset collaborative filtering algorithm and a preset content recommendation algorithm, the user preference prediction model performs matrix decomposition on each of the user behavior features to obtain the user preference information output by the user preference prediction model.

[0159] The application preloading device is also used for:

[0160] The data integration layer integrates and analyzes the current network environment data to obtain integrated network environment data.

[0161] The integrated network environment data and the user preference information are input into the model prediction layer for model evaluation, and the model prediction result output by the model prediction layer is obtained.

[0162] The model prediction results are input into the resource preloading layer to obtain the application recommendation list and resource preloading strategy output by the resource preloading layer.

[0163] The application preloading device is also used for:

[0164] Obtain user feedback and network changes;

[0165] Based on the user feedback and network changes, update the resource preloading strategy and the application recommendation list;

[0166] Based on the updated resource preloading strategy, the updated application recommendation list is preloaded using the preset backend cache library.

[0167] The application preloading device is also used for:

[0168] The application cache capacity and resource update frequency in the resource preloading strategy are dynamically updated based on the network changes.

[0169] The user feedback information is analyzed to obtain real-time analysis results;

[0170] Based on the real-time analysis results, the application priorities in the application recommendation list are updated.

[0171] The application preloading device is also used for:

[0172] Determine the app ranking in the updated app recommendation list;

[0173] Based on the cache time, cache size, and cache expiration policy in the updated resource preloading strategy, the applications in the updated application recommendation list are loaded into the preset backend cache library according to the application sorting.

[0174] The application preloading device is also used for:

[0175] Read the application loading request of the target user;

[0176] The application loading request is matched with the preset backend cache library to determine whether the match is successful.

[0177] If so, the application data corresponding to the application loading request is read from the preset backend cache library, and the application data is returned to the target user in response to the user request.

[0178] The application preloading device provided in this application, employing the application preloading method in the above embodiments, can solve the technical problems described in the background section. Compared with the prior art, the beneficial effects of the application preloading device provided in this application are the same as those of the application preloading method provided in the above embodiments, and other technical features in the application preloading device are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0179] This application provides an application preloading device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, which are executed by the at least one processor to enable the at least one processor to perform the application preloading method in Embodiment 1 above.

[0180] The following is for reference. Figure 10 The diagram illustrates a structural schematic of an application preloading device suitable for implementing embodiments of this application. The application preloading device in embodiments of this application may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Description), PMPs (Portable Media Players), in-vehicle terminals (e.g., in-vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. Figure 10 The application preloading device shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of this application.

[0181] like Figure 10As shown, the application preloading device may include a processing unit 1001 (e.g., a central processing unit, a graphics processing unit, etc.) that can perform various appropriate actions and processes according to a program stored in the read-only memory 1002 or a program loaded from the storage device 1003 into the random access memory 1004. The random access memory 1004 also stores various programs and data required for the operation of the application preloading device. The processing unit 1001, the read-only memory 1002, and the random access memory 1004 are interconnected via a bus 1005. An input / output interface 1006 is also connected to the bus. Typically, the following systems can be connected to the input / output interface 1006: input devices 1007 including, for example, touchscreens, touchpads, keyboards, mice, image sensors, microphones, accelerometers, gyroscopes, etc.; output devices 1008 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 1003 including, for example, magnetic tapes, hard disks, etc.; and communication devices 1009. Communication device 1009 allows the application pre-loaded device to communicate wirelessly or wiredly with other devices to exchange data. While the figure shows application pre-loaded devices with various systems, it should be understood that implementation or possession of all the systems shown is not required. More or fewer systems may be implemented alternatively.

[0182] Specifically, according to the embodiments disclosed in this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from read-only memory 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this application.

[0183] The application preloading device provided in this application, employing the application preloading method described in the above embodiments, can solve the technical problems mentioned in the background section. Compared with the prior art, the beneficial effects of the application preloading device provided in this application are the same as those of the application preloading method provided in the above embodiments, and other technical features of the application preloading device are the same as those disclosed in the previous embodiment method, and will not be repeated here.

[0184] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0185] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0186] This application provides a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon, the computer-readable program instructions being used to execute the application preloading method described in the above embodiments.

[0187] The computer-readable storage medium provided in this application may be, for example, a USB flash drive, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems or devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0188] The aforementioned computer-readable storage medium may be included in the application preload device; or it may exist independently and not be assembled into the application preload device.

[0189] The aforementioned computer-readable storage medium carries one or more programs that, when executed by the application preloading device, cause the application preloading device to:

[0190] Acquire user behavior data and current network environment data of the target user;

[0191] The user behavior data is input into the user preference prediction model for preference analysis to obtain the user preference information output by the user preference prediction model.

[0192] The user preference information and the current network environment data are input into the application recommendation engine for recommendation analysis, and the application recommendation list and resource preloading strategy output by the application recommendation engine are obtained.

[0193] Based on the resource preloading strategy, the application recommendation list is preloaded using a preset backend cache library.

[0194] Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a Local Area Network (LAN) or a Wide Area Network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0195] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0196] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0197] The readable storage medium provided in this application is a computer-readable storage medium that stores computer-readable program instructions (i.e., a computer program) for executing the above-described application preloading method, and is capable of solving the technical problems described in the background art. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as the beneficial effects of the application preloading method provided in the above embodiments, and will not be repeated here.

[0198] This application provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the application preloading method described above.

[0199] The computer program product provided in this application can solve the technical problems described in the background section. Compared with the prior art, the beneficial effects of the computer program product provided in the embodiments of this application are the same as the beneficial effects of the application preloading method provided in the above embodiments, and will not be repeated here.

[0200] The above description is only a part of the embodiments of this application and does not limit the patent scope of this application. All equivalent structural transformations made under the technical concept of this application and using the contents of the specification and drawings of this application, or direct / indirect applications in other related technical fields, are included in the patent protection scope of this application.

Claims

1. An application preloading method, characterized in that, include: Acquire user behavior data and current network environment data of the target user; The user behavior data is input into the user preference prediction model for preference analysis to obtain the user preference information output by the user preference prediction model. The user preference information and the current network environment data are input into the application recommendation engine for recommendation analysis, and the application recommendation list and resource preloading strategy output by the application recommendation engine are obtained. Based on the resource preloading strategy, the application recommendation list is preloaded using a preset backend cache library.

2. The application preloading method as described in claim 1, characterized in that, The step of inputting the user behavior data into a user preference prediction model for preference analysis to obtain user preference information output by the user preference prediction model includes: Feature extraction is performed on the user behavior data to obtain several user behavior features; Based on a preset collaborative filtering algorithm and a preset content recommendation algorithm, the user preference prediction model performs matrix decomposition on each of the user behavior features to obtain the user preference information output by the user preference prediction model.

3. The application preloading method as described in claim 1, characterized in that, The application recommendation engine includes a data integration layer, a model prediction layer, and a resource preloading layer. The step of inputting the user preference information and the current network environment data into the application recommendation engine for recommendation analysis, and obtaining the application recommendation list and resource preloading strategy output by the application recommendation engine, includes: The data integration layer integrates and analyzes the current network environment data to obtain integrated network environment data. The integrated network environment data and the user preference information are input into the model prediction layer for model evaluation, and the model prediction result output by the model prediction layer is obtained. The model prediction results are input into the resource preloading layer to obtain the application recommendation list and resource preloading strategy output by the resource preloading layer.

4. The application preloading method as described in claim 1, characterized in that, The step of preloading applications in the application recommendation list based on the resource preloading strategy and using a preset backend cache library includes: Obtain user feedback and network changes; Based on the user feedback and network changes, update the resource preloading strategy and the application recommendation list; Based on the updated resource preloading strategy, the updated application recommendation list is preloaded using the preset backend cache library.

5. The application preloading method as described in claim 4, characterized in that, The step of updating the resource preloading strategy and the application recommendation list based on the user feedback information and network changes includes: The application cache capacity and resource update frequency in the resource preloading strategy are dynamically updated based on the network changes. The user feedback information is analyzed to obtain real-time analysis results; Based on the real-time analysis results, the application priorities in the application recommendation list are updated.

6. The application preloading method as described in claim 4, characterized in that, The step of preloading applications based on the updated resource preloading strategy and using the preset backend cache library to preload the updated application recommendation list includes: Determine the app ranking in the updated app recommendation list; Based on the cache time, cache size, and cache expiration policy in the updated resource preloading strategy, the applications in the updated application recommendation list are loaded into the preset backend cache library according to the application sorting.

7. The application preloading method as described in claim 1, characterized in that, After preloading the application recommendation list using a preset backend cache library based on the resource preloading strategy, the process further includes: Read the application loading request of the target user; The application loading request is matched with the preset backend cache library to determine whether the match is successful. If so, the application data corresponding to the application loading request is read from the preset backend cache library, and the application data is returned to the target user in response to the user request.

8. An application preloading device, characterized in that, include: The data acquisition module is used to acquire user behavior data of the target user and current network environment data; The preference prediction module is used to input the user behavior data into the user preference prediction model for preference analysis, and obtain the user preference information output by the user preference prediction model. The application recommendation module is used to input the user preference information and the current network environment data into the application recommendation engine for recommendation analysis, and obtain the application recommendation list and resource preloading strategy output by the application recommendation engine. The application loading module is used to preload the application recommendation list based on the resource preloading strategy and through a preset backend cache library.

9. An application preloading device, characterized in that, The application preloading device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the application preloading method as described in any one of claims 1 to 7.

10. A storage medium, characterized in that, The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, it implements the steps of the application preloading method as described in any one of claims 1 to 7.

Images