A block recommendation method and device, computer equipment and storage medium
By configuring operation blocks in the operation zone and training a neural network model, the problems of high product recommendation cost and low efficiency in existing technologies are solved, and efficient personalized recommendation and accurate user matching are achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- CHINA PING AN PROPERTY INSURANCE CO LTD
- Filing Date
- 2023-06-28
- Publication Date
- 2026-06-26
AI Technical Summary
Existing technologies suffer from high product recommendation costs, long development cycles, and low operational efficiency, while homepage browsing performance relies on manual optimization, which is also costly.
Create an operations zone and configure operations blocks. Train a neural network model by acquiring historical user preference data, build a recommendation model, and make personalized recommendations based on user feature data.
It improved development efficiency, reduced costs and time, achieved higher accuracy and efficiency in personalized recommendations, and reduced user churn and labor costs.
Smart Images
Figure CN116776000B_ABST
Abstract
Description
Technical Field
[0001] This application relates to the fields of artificial intelligence technology and fintech, and in particular to a block recommendation method, apparatus, computer equipment, and storage medium. Background Technology
[0002] Currently, various e-commerce platforms have become indispensable tools in daily consumption, and people are increasingly inclined to browse information or shop online. Driven by the high traffic on the homepage, more service providers hope to gain homepage exposure, leading to a growing demand for expanding their content pool (recommendation pool). However, during operation, it has been found that the process of adding new recommendation pool members involves many stakeholders, update scheduling is time-consuming, and repetitive development costs are high. In addition, the homepage browsing experience is generally achieved by the front-end server, requiring optimization of the front-end server's operating strategy to improve the homepage browsing experience and response speed. In this regard, operating platforms mainly rely on manual integration and maintenance to achieve page personalization, which is very costly. Summary of the Invention
[0003] The purpose of this application is to provide a block recommendation method, apparatus, computer device, and storage medium to solve the technical problems of high product recommendation cost, long development cycle, and low operating efficiency in related technologies.
[0004] To address the aforementioned technical problems, this application provides a block recommendation method, employing the following technical solution:
[0005] Create an operations zone and configure operations blocks within the operations zone, wherein the operations blocks contain operations content;
[0006] Obtain sample data of operational blocks based on historical user preferences. The sample data includes input data and expected output. The input data includes basic characteristic data and behavioral characteristic data of the historical users, as well as block performance data and content characteristic data of the operational blocks. The expected output is the operational block corresponding to the input data.
[0007] Construct a neural network model, train the neural network model based on the input data and the desired output until the model converges, and output the final model as the recommendation model;
[0008] Obtain the target user's basic characteristic data and target behavior characteristic data, and obtain the target block performance data and target content characteristic data based on the target basic characteristic data and the target behavior characteristic data;
[0009] The target basic feature data, the target behavioral feature data, the target block performance data, and the target content feature data are input into the recommendation model, and the target operational block is output.
[0010] Furthermore, the steps for constructing the neural network model include:
[0011] The number of input nodes in the input layer is determined based on the number of feature attributes of the input data, and the number of output nodes in the output layer is determined based on the desired output.
[0012] Construct at least one hidden layer, and obtain the number of hidden neurons in each hidden layer based on the number of input nodes and the number of output nodes;
[0013] A neural network model is constructed based on the number of input nodes, the number of hidden neurons, and the number of output nodes.
[0014] Furthermore, the step of training the neural network model based on the input data and the desired output until the model converges, and outputting the final model as the recommendation model, includes:
[0015] An input feature vector is constructed based on the input data, and the input feature vector is input into the neural network model to output the prediction result.
[0016] The error result is calculated based on the prediction result and the expected output;
[0017] When the error result is greater than the preset error, the model parameters are adjusted according to the error gradient descent method until the error result is less than or equal to the preset error, the model converges, and the final model parameters are output as the target parameters.
[0018] The neural network model is updated according to the target parameters to obtain the recommendation model.
[0019] Furthermore, after the step of configuring the operation block within the operation zone, the method further includes:
[0020] Create a carousel view within the operation block, and create a carousel display control and a carousel parameter object within the carousel view. The carousel object parameters are configured with carousel data for the carousel content.
[0021] Load the carousel parameter objects into the carousel parameter object data group;
[0022] Upon receiving a carousel request, obtain the object data group corresponding to the carousel content;
[0023] The carousel view retrieves the carousel parameter object from the object data group;
[0024] Based on the carousel data in the carousel parameter object, the carousel content is displayed in the carousel display control at preset time intervals.
[0025] Furthermore, after the step of outputting the target operating block, the following is also included:
[0026] The user characteristics of the target user are determined based on the target basic characteristic data and the target behavioral characteristic data;
[0027] Based on the user characteristics, user categories are determined, and operational blocks are matched with priority recommendations for the target user based on the user categories.
[0028] Furthermore, after the step of outputting the target operating block, the following is also included:
[0029] Create a new operation block in the operation zone as a new operation block, and push the new operation block to the user;
[0030] Obtain tag data for the newly added operational blocks within a preset time period. The tag data includes block identification information and block performance data for the newly added operational blocks.
[0031] Based on the tag data, the newly added operation block is scored according to the preset scoring rules to obtain the scoring result;
[0032] The newly added operational blocks will be displayed based on the scoring results.
[0033] Furthermore, after the step of outputting the target operating block, the following is also included:
[0034] The system acquires real-time behavioral characteristic data of the target user and updates the displayed operation block based on the real-time behavioral characteristic data.
[0035] To address the aforementioned technical problems, this application also provides a block recommendation device, which employs the following technical solution:
[0036] The configuration module is used to create an operation zone and configure operation blocks within the operation zone, wherein the operation blocks contain operation content;
[0037] The first acquisition module is used to acquire sample data of the operation blocks of historical user preferences. The sample data includes input data and expected output. The input data includes the basic feature data and behavioral feature data of the historical user, as well as the block performance data and content feature data of the operation block. The expected output is the operation block corresponding to the input data.
[0038] The training module is used to build a neural network model, train the neural network model based on the input data and the expected output until the model converges, and output the final model as the recommendation model.
[0039] The second acquisition module is used to acquire the target user's target basic feature data and target behavior feature data, and to obtain the target block performance data and target content feature data based on the target basic feature data and the target behavior feature data.
[0040] The calculation module is used to input the target basic feature data, the target behavior feature data, the target block performance data, and the target content feature data into the recommendation model, and output the target operation block.
[0041] To address the aforementioned technical problems, this application also provides a computer device that employs the following technical solution:
[0042] The computer device includes a memory and a processor, the memory storing computer-readable instructions, and the processor executing the computer-readable instructions to implement the steps of the block recommendation method as described above.
[0043] To address the aforementioned technical problems, this application also provides a computer-readable storage medium, employing the technical solution described below:
[0044] The computer-readable storage medium stores computer-readable instructions that, when executed by a processor, implement the steps of the block recommendation method as described above.
[0045] Compared with the prior art, this application has the following main advantages:
[0046] This application creates an operations zone and configures operations blocks within it. These blocks contain operations content, expanding the recommended content through this feature. This approach is highly efficient and effectively reduces development costs and time. Furthermore, by training a neural network model using user basic and behavioral data, as well as the block performance and content features of the operations blocks, the resulting recommendation model tightly integrates user profiles with the operations blocks. This allows for timely and accurate recommendations to target users, improving the precision of personalized recommendations, preventing user churn, and simultaneously increasing recommendation efficiency while reducing the manpower costs of recommendation operations. Attached Figure Description
[0047] To more clearly illustrate the solutions in this application, the accompanying drawings used in the description of the embodiments of this application will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0048] Figure 1 This is an exemplary system architecture diagram to which this application can be applied;
[0049] Figure 2 This is a flowchart of one embodiment of the block recommendation method according to this application;
[0050] Figure 3 This is a schematic diagram of one embodiment of the block recommendation device according to this application;
[0051] Figure 4 This is a schematic diagram of the structure of one embodiment of the computer device according to this application. Detailed Implementation
[0052] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the terminology used herein in the specification of the application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application; the terms "comprising" and "having," and any variations thereof, in the specification, claims, and foregoing drawings of this application, are intended to cover non-exclusive inclusion. The terms "first," "second," etc., in the specification, claims, or foregoing drawings of this application are used to distinguish different objects, not to describe a particular order.
[0053] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.
[0054] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings.
[0055] This application provides a block recommendation method, which involves artificial intelligence and can be applied to, for example... Figure 1In the system architecture 100 shown, the system architecture 100 may include terminal devices 101, 102, and 103, a network 104, and a server 105. The network 104 is used as a medium to provide a communication link between the terminal devices 101, 102, and 103 and the server 105. The network 104 may include various connection types, such as wired or wireless communication links or fiber optic cables, etc.
[0056] Users can use terminal devices 101, 102, and 103 to interact with server 105 via network 104 to receive or send messages, etc. Various communication client applications can be installed on terminal devices 101, 102, and 103, such as web browser applications, shopping applications, search applications, instant messaging tools, email clients, social media platform software, etc.
[0057] Terminal devices 101, 102, and 103 can be various electronic devices with displays and support web browsing, including but not limited to smartphones, tablets, e-book readers, MP3 players (Moving Picture Experts Group Audio Layer III), MP4 players (Moving Picture Experts Group Audio Layer IV), laptops, and desktop computers, etc.
[0058] Server 105 can be a server that provides various services, such as a backend server that supports the pages displayed on terminal devices 101, 102, and 103.
[0059] It should be noted that the block recommendation method provided in this application embodiment is generally executed by a server / terminal device, and correspondingly, the block recommendation device is generally set in the server / terminal device.
[0060] It should be understood that Figure 1 The number of terminal devices, networks, and servers shown is merely illustrative. Depending on implementation needs, any number of terminal devices, networks, and servers can be included.
[0061] Continue to refer to Figure 2 The flowchart illustrates an embodiment of the block recommendation method according to this application, including the following steps:
[0062] Step S201: Create an operations zone and configure operations blocks within the operations zone. The operations blocks contain operations content.
[0063] Create an operations zone on the homepage. The operations zone has a corresponding content pool containing all the operations content of the operations zone. Multiple operations blocks are configured according to the characteristics and attributes of the operations content. The operations content includes, but is not limited to, products, insurance, services, content education, agricultural assistance, live streaming, special events, local benefits, and car owner benefits. Correspondingly, the operations blocks include, but are not limited to, money-saving zones, discover good products, car owner benefits, local benefits, car owner rights, special events, new user gift packs, insurance services, and agricultural assistance and poverty alleviation.
[0064] Step S202: Obtain sample data of historical user preference operation blocks. The sample data includes input data and expected output. The input data includes basic characteristic data and behavioral characteristic data of historical users, as well as block performance data and content characteristic data of operation blocks. The expected output is the operation block corresponding to the input data.
[0065] In this embodiment, the input data includes two parts: user data and operational block data. The user data includes basic feature data and behavioral feature data, while the operational block data includes block performance data and content feature data.
[0066] In this embodiment, basic feature data, behavioral feature data, block performance data, and content feature data each have their own feature attributes, and corresponding operational blocks need to be recommended to users based on these feature attributes.
[0067] Among them, the basic characteristic data of users are the basic attribute information obtained when registering as a user of the platform. The characteristic attributes include, but are not limited to, user insurance group, user vehicle category group, user family attributes, user driving experience, user vehicle attributes, and user membership level.
[0068] Specifically, user insurance groups include those who have purchased car insurance, those who have purchased non-car insurance, those who have purchased car insurance but not non-car insurance, those who have purchased non-car insurance but not car insurance, and those who have not purchased any type of insurance; user vehicle type groups include owners of new energy vehicles, owners of gasoline vehicles, and owners of hybrid vehicles; user family attributes include married, unmarried, and married with children; user driving experience includes new car owners with 0-1 years of experience, car owners with 1-3 years of experience, and car owners with more than 3 years of experience; user vehicle type includes sedans, sports cars, SUVs, and family cars; and user membership levels include non-members, ordinary members, gold, platinum, diamond, and supreme members.
[0069] The characteristic attributes of behavioral data include, but are not limited to, user browsing behavior, user interaction behavior, user purchasing behavior, user service behavior, and user shopping preferences. Specifically, user browsing behavior includes click behavior and page dwell time; user interaction behavior includes positive behaviors such as liking, sharing, commenting, and saving, and negative behaviors such as disliking and returning goods; user purchasing behavior includes product type, service type, and insurance type; user service behavior includes car washing, refueling, maintenance, and parking payment; user shopping preferences include price sensitivity, brand sensitivity, service attitude sensitivity, and location sensitivity.
[0070] The characteristic attributes of block performance data include, but are not limited to, block exposure count, block click data, block direct conversion data, and block indirect conversion data; among them, block direct conversion data indicates that users directly purchase after viewing; and block indirect conversion data indicates that users purchase the product within 7 days after viewing.
[0071] Content feature data consists of factors for evaluating each product category, which can be obtained from the basic attributes and statistical data system of each product. Product categories include goods, services, insurance, and content. Goods attributes include price, promotions (sales, group buying, etc.), sales volume (sales within a preset time period), positive review rate, and shipping speed. Services attributes include positive review rate of service outlets, outlet address, type of services offered (e.g., tire replacement, repair, car wash), sales volume, and customer reviews. Insurance attributes include insurance type (e.g., car insurance, non-car insurance), premium amount, and target audience. Content attributes include basic content type (e.g., articles, videos, live streams, topics, posts, Q&A, etc.), publication time (e.g., latest, within 3 days, within 7 days, within 30 days, etc.), page views, timeliness (whether it's trending), quality (whether the content is high-quality, judged by human intervention and click-through rate after exposure), category (insurance content, car maintenance content, new car content, etc.), length, author type (original author, institutional author, etc.), and tags.
[0072] It should be emphasized that, in order to further ensure the privacy and security of the basic feature data, the aforementioned basic feature data can also be stored in a blockchain node.
[0073] The blockchain referred to in this application is a novel application model of computer technologies such as distributed data storage, peer-to-peer transmission, consensus mechanisms, and encryption algorithms. Essentially, a blockchain is a decentralized database, a chain of data blocks linked together using cryptographic methods. Each data block contains information about a batch of network transactions, used to verify the validity of the information (anti-counterfeiting) and generate the next block. A blockchain can include an underlying blockchain platform, a platform product service layer, and an application service layer.
[0074] Step S203: Construct a neural network model. Train the neural network model based on the input data and the desired output until the model converges. Output the final model as the recommendation model.
[0075] In this embodiment, the neural network model can be a backpropagation neural network model, which includes an input layer, at least one hidden layer and an output layer.
[0076] The sample data is normalized to obtain normalized sample data, and the BP neural network model is trained using the normalized sample data.
[0077] The feature attributes of each user's normalized input data are used as the input vector, and the operational blocks preferred by that user are used as the output vector. The BP neural network model is trained using the input vector and the output vector until the model converges, thus obtaining the recommendation model.
[0078] Backpropagation (BP) neural networks are a type of feedforward neural network that uses backpropagation of error, and are also one of the most widely used types of neural networks. BP neural networks are a learning algorithm that continuously learns the intrinsic relationships between samples during the iterative training process using the principle of backpropagation of error, and applies these relationships to subsequent predictions. Therefore, using BP neural network models to predict sample data for operational blocks has significant guiding significance for personalized operational block recommendations for users.
[0079] In some embodiments, the steps of constructing the neural network model described above include:
[0080] The number of input nodes in the input layer is determined based on the number of feature attributes of the input data, and the number of output nodes in the output layer is determined based on the expected output.
[0081] Construct at least one hidden layer, and obtain the number of hidden neurons in each hidden layer based on the number of input nodes and the number of output nodes;
[0082] A neural network model is constructed based on the number of input nodes, the number of hidden neurons, and the number of output nodes.
[0083] The output nodes of the output layer can be set according to the desired number of outputs.
[0084] In this embodiment, the input data of each user is combined into a set of input vectors:
[0085] P(i)=P i1 P i2 P i3 , ..., P in ];
[0086] Among them, P ik Let P represent the k-th feature attribute of user i, where k = 1, ..., n, and n is the number of feature attributes of the user. ik When P = 0, the user does not possess this characteristic attribute; ik When the value is 1, the user possesses this characteristic attribute.
[0087] For example, the user's input vector is P(1) = P 11 P 12 P 13 , ..., P i2 The expression "]" indicates that the number of feature attributes in the input data is 12, and the input layer has 12 input nodes; the output layer has 4 output nodes, representing the output of 4 recommended operational blocks; the number of nodes in the hidden layer is... Where n is the number of input nodes, m is the number of output nodes, and a is a constant between 1 and 10.
[0088] The activation function for the hidden layer can be constructed using the transig function, as shown in the following formula:
[0089]
[0090] Where l represents the number of hidden layers; A l This represents the output vector of the l-th hidden layer in a BP neural network model. When l = 1, A0 = P, where P is the input vector of the input layer in the BP neural network model; W l and B l These represent the weight vector and threshold vector of the l-th hidden layer in the BP neural network model, respectively.
[0091] The activation function for the output layer is the logsig function.
[0092] Initialize the weight vectors and threshold vectors of each layer of the neural network model, and set the thresholds for the basic parameters of the neural network.
[0093] A BP neural network model is constructed based on a given number of input nodes, hidden neurons, output nodes, activation functions of the hidden layers, and activation functions of the output layers.
[0094] The BP neural network model is trained using input data according to the thresholds of the basic neural network parameters, which include the maximum number of training iterations, the minimum error of the training objective, the learning rate, and the maximum number of failures.
[0095] In some alternative implementations, the steps described above—training a neural network model based on input data and desired output until the model converges, and outputting the final model as a recommendation model—include:
[0096] Construct an input feature vector based on the input data, input the input feature vector into the neural network model, and output the prediction result;
[0097] The error result is calculated based on the prediction results and the expected output.
[0098] When the error result is greater than the preset error, the model parameters are adjusted according to the error gradient descent method until the error result is less than or equal to the preset error, the model converges, and the final model parameters are output as the target parameters.
[0099] The neural network model is updated according to the target parameters to obtain the recommendation model.
[0100] In this embodiment, the neural network model includes a forward propagation process and a backward propagation process. The forward propagation process is as follows: the input layer receives input data and passes it to the neurons in the intermediate layers (each hidden layer). Each neuron performs data processing and transformation, and then passes it to the output layer through the last hidden layer for output. The backward propagation process is as follows: after forward propagation, the error is obtained by comparing the actual value (expected output) and the output value (predicted result). When the error is greater than the preset error, that is, when the actual output differs too much from the expected output, the backward propagation stage is entered. The error passes through the output layer and corrects the parameters of each layer (such as weights, thresholds, and learning rates) in reverse according to the error gradient descent method, such as stochastic gradient descent (SGD), and then reverses layer by layer to the hidden layer and input layer.
[0101] Through continuous forward and backward propagation, the output error is reduced to the preset error or the maximum number of training iterations is reached, i.e., the model converges. The final model parameters are then output as the target parameters, and the neural network model is updated according to the target parameters to obtain the recommendation model.
[0102] The weight vectors and threshold vectors obtained after training can ensure that the subsequent operation block recommendations can obtain relatively accurate recommendation results.
[0103] Step S204: Obtain the target user's basic feature data and target behavior feature data, and obtain the target block performance data and target content feature data based on the target user's basic feature data and target behavior feature data.
[0104] In this embodiment, based on the target basic feature data and target behavior feature data, operational block information is extracted, such as the operational block clicked by the user, the operational block where the purchased product is located, etc.; the corresponding operational block is matched as the target operational block according to the operational block information, and then the target block performance data and target content feature data of the target operational block are obtained.
[0105] Step S205: Input the target basic feature data, target behavior feature data, target block performance data, and target content feature data into the recommendation model, and output the target operation block.
[0106] The target's basic feature data, target's behavioral feature data, target's block performance data, and target's content feature data are input into the recommendation model. The recommendation model calculates the target operation blocks that match the target user and recommends the target operation blocks to the target user.
[0107] In this embodiment, recommending target operational blocks to users based on user data and operational block data can improve the accuracy of personalized recommendations.
[0108] This application creates an operations zone and configures operations blocks within it. These blocks contain operations content, expanding the recommended content through this feature. This approach is highly efficient and effectively reduces development costs and time. Furthermore, by training a neural network model using user basic and behavioral data, as well as the block performance and content features of the operations blocks, the resulting recommendation model tightly integrates user profiles with the operations blocks. This allows for timely and accurate recommendations to target users, improving the precision of personalized recommendations, preventing user churn, and simultaneously increasing recommendation efficiency while reducing the manpower costs of recommendation operations.
[0109] In some alternative implementations, the following steps are included after configuring the operation block within the operation zone:
[0110] Create a carousel view within the operation block, and create a carousel display control and a carousel parameter object within the carousel view. The carousel object parameters are configured with the carousel data for the carousel content.
[0111] Load the carousel parameter objects into the carousel parameter object data group;
[0112] Upon receiving a carousel request, retrieve the object data group corresponding to the carousel content;
[0113] Retrieve the carousel parameter object from the object data group using the carousel view;
[0114] Based on the carousel data in the carousel parameter object, the carousel content is displayed in the carousel display control at preset time intervals.
[0115] In this embodiment, the operations block is used to carry out segmented campaigns, local operational content, and initial release information, supporting scenario-based operations. A preset number of carousel views (banners) can be created within the operations block to rotate the operational content. For example, the entire operations zone supports 2 or 4 cards, with one card corresponding to one operations block. Each card contains two banner styles; for example, the banners for the car owner benefits and "Discover Good Deals" blocks should support carousel rotation. Carousel styles include, but are not limited to, regular carousel styles, card styles, marquees, and card overlays.
[0116] A carousel view can be a BannerView. The BannerView is used to display the carousel content, and all initialization and data refresh loading of the carousel are completed within this view.
[0117] The carousel parameter objects include the imageView control and the BanenrParam object. The imageView control is the carousel display control on the BannerView view, and the BanenrParam object is the carousel parameter object, which is the data object to be displayed in the carousel. For example, one BanenrParam object corresponds to one carousel image; if there are multiple carousel images, multiple BanenrParam objects need to be created.
[0118] In this embodiment, the carousel data of the carousel content is configured in the carousel parameter object, and the carousel parameter object is loaded into the carousel parameter object data group. The object data group is a collection of multiple carousel images, and each object corresponds to the carousel data of one carousel image. The carousel data includes, but is not limited to, the carousel image data source, title, jump link, carousel sliding style, and text display style.
[0119] In this embodiment, after determining the operation block recommended to the target user, while displaying the operation block to the target user, a carousel request corresponding to the operation block is sent to the carousel view. The object data group corresponding to the carousel content is obtained through the carousel view, and the carousel parameter object in the object data group is retrieved. According to the carousel data in the carousel parameter object, the carousel content is displayed in the carousel display control at a preset time interval.
[0120] This application uses a carousel format to display different operational content, which can increase the exposure and popularity of the operational content.
[0121] In some alternative implementations, the step of outputting the target operational block described above also includes:
[0122] Determine the user characteristics of the target user based on the target's basic characteristic data and target's behavioral characteristic data;
[0123] User categories are determined based on user characteristics, and operational blocks are matched with priority recommendations for target users based on user categories.
[0124] In this embodiment, after the recommendation model outputs the target operation block for recommendation, the recommendation results for the target user can be adjusted according to a preset optimization strategy, wherein the optimization strategy is to make targeted recommendations based on user category.
[0125] Specifically, user characteristics are determined based on the characteristic attributes of the target's basic characteristic data and target's behavioral characteristic data. User categories are then determined based on these user characteristics. User categories include, but are not limited to, new users, regular users, and members. When a user is identified as a new user, the new user zone is given priority.
[0126] In some embodiments, the platform may have activities that need to be launched recently, such as promotional activities, group buying activities, or flash sales, and may prioritize recommending the corresponding operational blocks.
[0127] In some embodiments, live stream content associated with the operational content in the target operational block is obtained from the live streaming subsystem on the server side. When a user is browsing the details page of the corresponding operational content, the live stream content can be directly dragged into the content container of the details page for playback, so as to better display it to the user.
[0128] This embodiment improves the accuracy of personalized recommendations and enhances user experience by adjusting the recommendation strategy based on user categories.
[0129] In some alternative implementations, the step of outputting the target operational block described above also includes:
[0130] Create a new operation block in the operation zone as a new operation block, and push the new operation block to the user;
[0131] Obtain the tag data of the newly added operating blocks within a preset time period. The tag data includes the block identification information and block performance data of the newly added operating blocks.
[0132] Based on the tag data, the newly added operation blocks are scored according to the preset scoring rules to obtain the scoring results;
[0133] The newly added operational blocks will be displayed based on the scoring results.
[0134] In this embodiment, when a new product is launched that needs to be recommended, an operational block for the new product is configured as a new operational block in the content pool of the operational zone. This new operational block is then prioritized and pushed to users. After a period of time, tag data for the new operational block within a preset time frame (e.g., 10 days, half a month, or one month) is obtained. The tag data includes the block identifier information and block performance data. The block identifier information includes the block ID and block name, while the block performance data includes the number of times the block is exposed and the number of clicks. Based on the tag data, the new operational block is scored according to a preset scoring rule. The new operational block pushed to users on the homepage is then replaced based on the scoring results. For example, if a block is exposed three times without any user clicks, the click-through rate is 0%, and it will be replaced by a new block.
[0135] In some embodiments, the operation block can be closed by a user clicking it.
[0136] This embodiment scores newly added operational blocks and determines whether to display them to users based on the scoring results, thus providing users with better personalized recommendations.
[0137] In some alternative implementations, the step of outputting the target operational block described above also includes:
[0138] Real-time acquisition of target user behavior data, and updating of the displayed operation blocks based on the real-time behavior data.
[0139] After a target user enters the platform, real-time behavioral characteristic data of the target user is collected, such as the target user's click operations and browsing trajectory. This real-time behavioral characteristic data is transmitted to the recommendation model or big data center, which calculates the operational blocks that match the target user and updates the homepage.
[0140] In this embodiment, the displayed operation blocks are dynamically adjusted by using real-time user behavior data, which ensures the accuracy of the recommended operation blocks, realizes automated recommendation and operation maintenance of operation blocks, further reduces the workload of operations, and lowers labor costs.
[0141] This application can be used in a wide variety of general-purpose or special-purpose computer system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices. This application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform specific tasks or implement specific abstract data types. This application can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.
[0142] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by instructing related hardware with computer-readable instructions. These computer-readable instructions can be stored in a computer-readable storage medium. When executed, the program can include the processes of the embodiments of the above methods. The aforementioned storage medium can be a non-volatile storage medium such as a magnetic disk, optical disk, or read-only memory (ROM), or random access memory (RAM).
[0143] It should be understood that although the steps in the flowcharts of the accompanying figures are shown sequentially as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the accompanying figures may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.
[0144] Further reference Figure 3 As a response to the above Figure 2 To implement the method shown, this application provides an embodiment of a block recommendation device, which is similar to... Figure 2 Corresponding to the method embodiments shown, this device can be specifically applied to various electronic devices.
[0145] like Figure 3As shown, the block recommendation device 300 described in this embodiment includes: a configuration module 301, a first acquisition module 302, a training module 303, a second acquisition module 304, and a calculation module 305. Wherein:
[0146] The configuration module 301 is used to create an operation zone and configure operation blocks within the operation zone, wherein the operation blocks contain operation content;
[0147] The first acquisition module 302 is used to acquire sample data of the operation blocks of historical user preferences. The sample data includes input data and expected output. The input data includes the basic feature data and behavioral feature data of the historical user, as well as the block performance data and content feature data of the operation block. The expected output is the operation block corresponding to the input data.
[0148] The training module 303 is used to build a neural network model, train the neural network model based on the input data and the expected output until the model converges, and output the final model as the recommendation model.
[0149] The second acquisition module 304 is used to acquire the target user's target basic feature data and target behavior feature data, and to obtain the target block performance data and target content feature data based on the target basic feature data and the target behavior feature data.
[0150] The calculation module 305 is used to input the target basic feature data, the target behavior feature data, the target block performance data and the target content feature data into the recommendation model, and output the target operation block.
[0151] It should be emphasized that, in order to further ensure the privacy and security of the basic feature data, the aforementioned basic feature data can also be stored in a blockchain node.
[0152] Based on the aforementioned block recommendation device, by creating an operation zone and configuring operation blocks within that zone, the recommended products can be expanded by configuring operation blocks within the operation zone. This approach achieves high development efficiency and effectively reduces development costs and time. By training a neural network model using user basic characteristic data, behavioral characteristic data, and block performance and content characteristic data from the operation blocks, the resulting recommendation model can tightly integrate user profiles with operation blocks. This allows for timely and accurate product recommendations to target users, improving the precision of personalized recommendations, preventing user churn, and simultaneously increasing recommendation efficiency while reducing the manpower costs of recommendation operations.
[0153] In some optional implementations of this embodiment, the training module 303 includes a construction submodule, which is used for:
[0154] The number of input nodes in the input layer is determined based on the number of feature attributes of the input data, and the number of output nodes in the output layer is determined based on the desired output.
[0155] Construct at least one hidden layer, and obtain the number of hidden neurons in each hidden layer based on the number of input nodes and the number of output nodes;
[0156] A neural network model is constructed based on the number of input nodes, the number of hidden neurons, and the number of output nodes.
[0157] In this embodiment, the BP neural network model is used to predict the sample data of the operation block, which has important guiding significance for personalized operation block recommendations for users.
[0158] In this embodiment, the training module 303 includes a forward propagation submodule, an error calculation submodule, an adjustment submodule, and an update submodule, wherein:
[0159] The forward propagation submodule is used to construct an input feature vector based on the input data, input the input feature vector into the neural network model, and output the prediction result.
[0160] The error calculation submodule is used to calculate the error result based on the prediction result and the expected output;
[0161] The adjustment submodule is used to adjust the model parameters according to the error gradient descent method when the error result is greater than the preset error, until the error result is less than or equal to the preset error, the model converges, and the final model parameters are output as the target parameters.
[0162] The update submodule is used to update the neural network model according to the target parameters to obtain the recommendation model.
[0163] The recommendation model obtained through training in this embodiment can ensure that subsequent operational block recommendations will yield relatively accurate results.
[0164] In some alternative implementations, the block recommendation device 300 also includes a carousel module, which is used for:
[0165] Create a carousel view within the operation block, and create a carousel display control and a carousel parameter object within the carousel view. The carousel object parameters are configured with carousel data for the carousel content.
[0166] Load the carousel parameter objects into the carousel parameter object data group;
[0167] Upon receiving a carousel request, obtain the object data group corresponding to the carousel content;
[0168] The carousel view retrieves the carousel parameter object from the object data group;
[0169] Based on the carousel data in the carousel parameter object, the carousel content is displayed in the carousel display control at preset time intervals.
[0170] Displaying different operational content through a carousel can increase the exposure and popularity of that content.
[0171] In some alternative implementations, the block recommendation device 300 further includes a priority recommendation module, which is used to determine the user characteristics of the target user based on the target basic feature data and the target behavioral feature data; determine the user category based on the user characteristics; and match the target user with priority recommended operating blocks based on the user category.
[0172] Adjusting recommendation strategies by user categories can improve the accuracy of personalized recommendations and enhance user experience.
[0173] In some selected implementations, the block recommendation device 300 also includes a new module, which is used for:
[0174] Create a new operation block in the operation zone as a new operation block, and push the new operation block to the user;
[0175] Obtain tag data for the newly added operational blocks within a preset time period. The tag data includes block identification information and block performance data for the newly added operational blocks.
[0176] Based on the tag data, the newly added operation block is scored according to the preset scoring rules to obtain the scoring result;
[0177] The newly added operational blocks will be displayed based on the scoring results.
[0178] This embodiment scores newly added operational blocks and determines whether to display them to users based on the scoring results, thus providing users with better personalized recommendations.
[0179] In some selected implementations, the block recommendation device 300 also includes a real-time update module, which is used to acquire the real-time behavioral feature data of the target user in real time and update the displayed operation blocks based on the real-time behavioral feature data.
[0180] By dynamically adjusting the displayed operational blocks based on users' real-time behavioral data, the accuracy of the recommended operational blocks can be guaranteed. This enables automated recommendation and operation maintenance of operational blocks, further reducing the workload of operations and lowering labor costs.
[0181] To address the aforementioned technical problems, embodiments of this application also provide a computer device. Please refer to [link / reference needed]. Figure 4 , Figure 4 This is a basic structural block diagram of the computer device in this embodiment.
[0182] The computer device 4 includes a memory 41, a processor 42, and a network interface 43 that are interconnected via a system bus. It should be noted that only the computer device 4 with components 41-43 is shown in the figure; however, it should be understood that it is not required to implement all the shown components, and more or fewer components can be implemented alternatively. Those skilled in the art will understand that the computer device described here is a device capable of automatically performing numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes, but is not limited to, microprocessors, application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), digital signal processors (DSPs), embedded devices, etc.
[0183] The computer device can be a desktop computer, laptop, handheld computer, or cloud server, etc. The computer device can interact with the user via a keyboard, mouse, remote control, touchpad, or voice control.
[0184] The memory 41 includes at least one type of readable storage medium, including flash memory, hard disk, multimedia card, card-type memory (e.g., SD or DX memory), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the memory 41 may be an internal storage unit of the computer device 4, such as the hard disk or memory of the computer device 4. In other embodiments, the memory 41 may also be an external storage device of the computer device 4, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on the computer device 4. Of course, the memory 41 may also include both the internal storage unit and its external storage device of the computer device 4. In this embodiment, the memory 41 is typically used to store the operating system and various application software installed on the computer device 4, such as computer-readable instructions for block recommendation methods. In addition, the memory 41 can also be used to temporarily store various types of data that have been output or will be output.
[0185] In some embodiments, the processor 42 may be a central processing unit (CPU), controller, microcontroller, microprocessor, or other data processing chip. The processor 42 is typically used to control the overall operation of the computer device 4. In this embodiment, the processor 42 is used to execute computer-readable instructions stored in the memory 41 or to process data, such as executing computer-readable instructions for the block recommendation method.
[0186] The network interface 43 may include a wireless network interface or a wired network interface, which is typically used to establish communication connections between the computer device 4 and other electronic devices.
[0187] This embodiment implements the steps of the block recommendation method described above by executing computer-readable instructions stored in memory through a processor. It creates an operational zone and configures operational blocks within that zone, each containing operational content. This expands the recommended content by configuring operational blocks within the operational zone, resulting in high development efficiency and effectively reducing development costs and time. By training a neural network model using user basic characteristic data, behavioral characteristic data, and the block performance and content characteristic data of the operational blocks, the resulting recommendation model tightly integrates user profiles and operational blocks. This allows for timely and accurate recommendations to target users, improving the precision of personalized recommendations, preventing user churn, and simultaneously increasing recommendation efficiency while reducing the manpower costs of recommendation operations.
[0188] This application also provides another implementation method, namely, a computer-readable storage medium storing computer-readable instructions that can be executed by at least one processor to perform the steps of the block recommendation method described above. This method involves creating an operational zone and configuring operational blocks within that zone. These operational blocks contain operational content, and the recommended content is expanded by configuring operational blocks within the operational zone. This approach achieves high development efficiency and effectively reduces development costs and time. Furthermore, by training a neural network model using user basic characteristic data and behavioral characteristic data, as well as block performance data and content characteristic data from operational blocks, the resulting recommendation model can tightly integrate user profiles and operational blocks. This allows for timely and accurate recommendations to target users, improving the accuracy of personalized recommendations, preventing user churn, and simultaneously increasing recommendation efficiency while reducing the manpower costs of recommendation operations.
[0189] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of this application.
[0190] Obviously, the embodiments described above are only some embodiments of this application, not all embodiments. The accompanying drawings show preferred embodiments of this application, but do not limit the patent scope of this application. This application can be implemented in many different forms; rather, the purpose of providing these embodiments is to provide a more thorough and comprehensive understanding of the disclosure of this application. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing specific embodiments, or make equivalent substitutions for some of the technical features. Any equivalent structures made using the content of this application's specification and drawings, directly or indirectly applied to other related technical fields, are similarly within the scope of patent protection of this application.
Claims
1. A block recommendation method, characterized in that, Includes the following steps: Create an operations zone and configure operations blocks within the operations zone, wherein the operations blocks contain operations content; Obtain sample data of operational blocks based on historical user preferences. The sample data includes input data and expected output. The input data includes basic characteristic data and behavioral characteristic data of the historical users, as well as block performance data and content characteristic data of the operational blocks. The expected output is the operational block corresponding to the input data. The block performance data includes block exposure counts, block click data, block direct conversion data, and block indirect conversion data. Construct a neural network model, train the neural network model based on the input data and the desired output until the model converges, and output the final model as the recommendation model; Obtain the target user's basic characteristic data and target behavior characteristic data, and obtain the target block performance data and target content characteristic data based on the target basic characteristic data and the target behavior characteristic data; The target basic feature data, the target behavioral feature data, the target block performance data, and the target content feature data are input into the recommendation model, and the target operational block is output.
2. The block recommendation method according to claim 1, characterized in that, The steps for constructing the neural network model include: The number of input nodes in the input layer is determined based on the number of feature attributes of the input data, and the number of output nodes in the output layer is determined based on the desired output. Construct at least one hidden layer, and obtain the number of hidden neurons in each hidden layer based on the number of input nodes and the number of output nodes; A neural network model is constructed based on the number of input nodes, the number of hidden neurons, and the number of output nodes.
3. The block recommendation method according to claim 1 or 2, characterized in that, The steps of training the neural network model based on the input data and the expected output until the model converges, and outputting the final model as the recommendation model, include: An input feature vector is constructed based on the input data, and the input feature vector is input into the neural network model to output the prediction result. The error result is calculated based on the prediction result and the expected output; When the error result is greater than the preset error, the model parameters are adjusted according to the error gradient descent method until the error result is less than or equal to the preset error, the model converges, and the final model parameters are output as the target parameters. The neural network model is updated according to the target parameters to obtain the recommendation model.
4. The block recommendation method according to claim 1, characterized in that, Following the step of configuring the operation block within the operation zone, the method further includes: Create a carousel view within the operation block, and create a carousel display control and a carousel parameter object within the carousel view. The carousel object parameters are configured with carousel data for the carousel content. Load the carousel parameter objects into the carousel parameter object data group; Upon receiving a carousel request, obtain the object data group corresponding to the carousel content; The carousel view retrieves the carousel parameter object from the object data group; Based on the carousel data in the carousel parameter object, the carousel content is displayed in the carousel display control at preset time intervals.
5. The block recommendation method according to claim 1, characterized in that, The step of outputting the target operating block also includes: The user characteristics of the target user are determined based on the target basic characteristic data and the target behavioral characteristic data; Based on the user characteristics, user categories are determined, and operational blocks are matched with priority recommendations for the target user based on the user categories.
6. The block recommendation method according to claim 1, characterized in that, The step of outputting the target operating block also includes: Create a new operation block in the operation zone as a new operation block, and push the new operation block to the user; Obtain tag data for the newly added operational blocks within a preset time period. The tag data includes block identification information and block performance data for the newly added operational blocks. Based on the tag data, the newly added operation block is scored according to the preset scoring rules to obtain the scoring result; The newly added operational blocks will be displayed based on the scoring results.
7. The block recommendation method according to claim 1, characterized in that, The step of outputting the target operating block also includes: The system acquires real-time behavioral characteristic data of the target user and updates the displayed operation block based on the real-time behavioral characteristic data.
8. A block recommendation device, characterized in that, include: The configuration module is used to create an operation zone and configure operation blocks within the operation zone, wherein the operation blocks contain operation content; The first acquisition module is used to acquire sample data of operational blocks based on historical user preferences. The sample data includes input data and expected output. The input data includes basic characteristic data and behavioral characteristic data of the historical user, as well as block performance data and content characteristic data of the operational block. The expected output is the operational block corresponding to the input data. The block performance data includes block exposure count, block click data, block direct conversion data, and block indirect conversion data. The training module is used to build a neural network model, train the neural network model based on the input data and the expected output until the model converges, and output the final model as the recommendation model. The second acquisition module is used to acquire the target user's target basic feature data and target behavior feature data, and to obtain the target block performance data and target content feature data based on the target basic feature data and the target behavior feature data. The calculation module is used to input the target basic feature data, the target behavior feature data, the target block performance data, and the target content feature data into the recommendation model, and output the target operation block.
9. A computer device comprising a memory and a processor, the memory storing computer-readable instructions, wherein the processor, when executing the computer-readable instructions, implements the steps of the block recommendation method as described in any one of claims 1 to 7.
10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-readable instructions that, when executed by a processor, implement the steps of the block recommendation method as described in any one of claims 1 to 7.