Material recommendation method and device, electronic equipment and storage medium
By obtaining and configuring user characteristics and material characteristics, simulating recommendation needs, and updating the material library characteristics of the recommendation system, the problems of high resource occupation and low efficiency of existing recommendation systems are solved, and flexible and efficient material recommendations are achieved.
Patent Information
- Application Number
- CN202411208230.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-30
- Publication Date
- 2024-12-06
AI Technical Summary
When existing recommendation systems use a lot of resources and operate in low efficiency when they require scenarios such as online problem positioning, new function experience and recommendation results traceability, they need to frequently modify the recommendation logic, resulting in complex links and waste of resources.
By obtaining user characteristics and material characteristics, building user portraits and material characteristics, simulating recommendation requirements based on configuration feature data, updating preset material library features, realizing recommendation processing, and reducing redevelopment of the original logic.
It simplifies the configuration and optimization of the recommendation system, reduces resource usage, improves operating efficiency, and facilitates material recommendations that adapt to different scenarios and popular trends.
Smart Images

Figure CN119089016A_ABST
Abstract
Description
Technical Field
[0001] The present disclosure relates to but is not limited to the field of Internet technology, and in particular to a material recommendation method and device, an electronic device, and a storage medium. Background Art
[0002] The recommendation system is an information filtering system that can recommend suitable materials to users through layer-by-layer screening from a large number of candidate sets. Before a material is exposed to the user, it will go through multiple processes such as screening and storage, multi-channel recall, rough sorting, fine sorting, re-sorting and / or strategy adjustment. In each process, models and strategies may coexist and influence and overlap each other.
[0003] In some demand scenarios, such as online problem location, new function experience and / or recommendation result tracing, it is often necessary to modify the entire process or multiple modules in the recommendation system, resulting in high resource usage and low operating efficiency. Summary of the invention
[0004] In view of this, the embodiments of the present disclosure provide a material recommendation method and device, an electronic device and a storage medium, which can easily simulate any recommendation requirements through configuration.
[0005] The technical solution of the present disclosure is achieved as follows:
[0006] In a first aspect, an embodiment of the present disclosure provides a material recommendation method, comprising:
[0007] Acquire characteristic data configured for a target user; wherein the characteristic data includes user characteristics and / or material characteristics;
[0008] In response to detecting a recommendation request for the target user, based on the configured feature data and the initial material features of the preset material library, determining a user profile of the target user and target material features of the preset material library; wherein the configured user features are used to construct the user profile of the target user, and the configured material features are used to update the initial material features of the preset material library to generate the target material features;
[0009] The recommendation request is processed based on the user portrait of the target user and the characteristics of the target material to obtain a target material recommended for the target user.
[0010] In some embodiments, in response to detecting a recommendation request for the target user, determining a user profile of the target user and target material characteristics of the preset material library based on the configured feature data and initial material characteristics of the preset material library includes:
[0011] In response to detecting a recommendation request for the target user in an offline simulated recommendation environment, and the configured feature data includes the user features and the material features, using the configured user features to replace the actual user features of the target user and constructing a user profile of the target user, and updating the configured material features to the initial material features to obtain the target material features;
[0012] In response to detecting a recommendation request for the target user in an offline simulated recommendation environment, and the configured feature data includes the user features, the configured user features are used to replace the actual user features of the target user and a user profile of the target user is constructed, and the initial material features are used as the target material features.
[0013] In some embodiments, in response to detecting a recommendation request for the target user, determining a user profile of the target user and target material characteristics of the preset material library based on the configured feature data and initial material characteristics of the preset material library includes:
[0014] In response to detecting a recommendation request for the target user in an online recommendation environment, and the configured feature data includes the user features and the material features, constructing a user profile of the target user based on original user features of the target user obtained online and the configured user features, and updating the configured material features to the initial material features to obtain the target material features;
[0015] In response to detecting a recommendation request for the target user in an online recommendation environment, and the configured feature data includes the user features, constructing a user profile of the target user based on original user features of the target user obtained online and the configured user features, and using the initial material features as the target material features;
[0016] In response to detecting a recommendation request for the target user in an online recommendation environment, and the configured feature data includes the material features, a user profile of the target user is constructed based on the original user features of the target user obtained online, and the configured material features are updated to the initial material features to obtain the target material features.
[0017] In some embodiments, the determining the user portrait of the target user and the target material characteristics of the preset material library based on the configured feature data and the initial material characteristics of the preset material library includes:
[0018] The configured feature data is adapted based on the recommended processing requirements to obtain adapted feature data; the adapted feature data includes adapted user features and / or material features;
[0019] Based on the adapted feature data and the initial material features of the preset material library, the user portrait of the target user and the target material features of the preset material library are determined.
[0020] In some embodiments, the target user is a user with feature data configuration permission.
[0021] In some embodiments, the method further comprises:
[0022] In the process of performing recommendation processing on the recommendation request based on the user portrait of the target user and the target material characteristics, saving the material screening results of the intermediate stage; wherein the intermediate stage includes at least one of the following: recall, rough sorting, fine sorting, and re-sorting;
[0023] The page displays the material screening results of the intermediate stage.
[0024] In some embodiments, the page displays the material screening results of the intermediate stage, including:
[0025] In response to detecting an instruction to display the processing result of any intermediate stage, the page displays the material screening result of the intermediate stage indicated by the instruction.
[0026] In a second aspect, an embodiment of the present disclosure provides a material recommendation device, including:
[0027] An acquisition module, configured to acquire feature data configured for a target user; wherein the feature data includes user features and / or material features;
[0028] a determination module configured to, in response to detecting a recommendation request for the target user, determine a user profile of the target user and target material features of the preset material library based on the configured feature data and the initial material features of the preset material library; wherein the configured user features are used to construct the user profile of the target user, and the configured material features are used to update the initial material features of the preset material library to generate the target material features;
[0029] The recommendation module is configured to perform recommendation processing on the recommendation request based on the user portrait of the target user and the characteristics of the target material to obtain the target material recommended for the target user.
[0030] In some embodiments, the determination module is further configured to: in response to detecting a recommendation request for the target user in an offline simulated recommendation environment, and the configured feature data includes the user features and the material features, use the configured user features to replace the actual user features of the target user and construct a user profile of the target user, and update the configured material features to the initial material features to obtain the target material features; in response to detecting a recommendation request for the target user in an offline simulated recommendation environment, and the configured feature data includes the user features, use the configured user features to replace the actual user features of the target user and construct a user profile of the target user, and use the initial material features as the target material features.
[0031] In some embodiments, the determination module is further configured to: in response to detecting a recommendation request for the target user in an online recommendation environment, and the configured feature data includes the user features and the material features, construct a user profile of the target user based on the original user features of the target user obtained online and the configured user features, and update the configured material features to the initial material features to obtain the target material features; in response to detecting a recommendation request for the target user in an online recommendation environment, and the configured feature data includes the user features, construct a user profile of the target user based on the original user features of the target user obtained online and the configured user features, and use the initial material features as the target material features; in response to detecting a recommendation request for the target user in an online recommendation environment, and the configured feature data includes the material features, construct a user profile of the target user based on the original user features of the target user obtained online, and update the configured material features to the initial material features to obtain the target material features.
[0032] In some embodiments, the determination module is further configured to: adapt the configured feature data based on the recommended processing requirements to obtain adapted feature data; the adapted feature data includes adapted user characteristics and / or material characteristics; based on the adapted feature data and the initial material characteristics of the preset material library, determine the user portrait of the target user and the target material characteristics of the preset material library.
[0033] In some embodiments, the target user is a user with feature data configuration permission.
[0034] In some embodiments, the apparatus further comprises:
[0035] A saving module is configured to save the material screening results of the intermediate stage in the process of performing recommendation processing on the recommendation request based on the user portrait of the target user and the target material characteristics; wherein the intermediate stage includes at least one of the following: recall, rough sorting, fine sorting, and re-sorting;
[0036] The display module is configured to display the material screening results of the intermediate stage on a page.
[0037] In some embodiments, the display module is further configured to: in response to detecting an instruction to display the processing results of any intermediate stage, display the material screening results of the intermediate stage indicated by the instruction on the page.
[0038] In a third aspect, an embodiment of the present disclosure provides an electronic device, including:
[0039] A memory for storing executable instructions;
[0040] The processor is used to implement the material recommendation method of the embodiment of the present disclosure when executing the executable instructions stored in the memory.
[0041] In a fourth aspect, an embodiment of the present disclosure provides a computer storage medium, wherein the computer storage instructions store a computer program or executable instructions, and when the computer program or executable instructions are executed by a processor, the material recommendation method of the embodiment of the present disclosure is implemented.
[0042] In a fifth aspect, an embodiment of the present disclosure provides a computer program product, which implements the material recommendation method of the embodiment of the present disclosure when the computer program or instructions are executed by a processor.
[0043] In the embodiments of the present disclosure, the configuration of user features and / or material features is supported. Since both user features and material features are the features most directly related to the recommendation results, the embodiments of the present disclosure optimize the problem from the underlying data dimension, making it convenient to simulate any recommendation requirements (such as scene reproduction or new function experience) through configuration without redeveloping the recommendation logic. This can reduce the problems of high resource usage and low operating efficiency caused by redeveloping the original recommendation logic. BRIEF DESCRIPTION OF THE DRAWINGS
[0044] Figure 1 A flow chart of a material recommendation method provided in an embodiment of the present disclosure.
[0045] Figure 2 This is an example diagram of page configuration feature data in an embodiment of the present disclosure.
[0046] Figure 3 This is an example diagram of a page display in an embodiment of the present disclosure.
[0047] Figure 4Schematic diagram of the material recommendation method in the embodiment of the present disclosure.
[0048] Figure 5 A material recommendation device diagram is provided for an embodiment of the present disclosure.
[0049] Figure 6 A schematic diagram of a hardware entity of an electronic device in an embodiment of the present disclosure. DETAILED DESCRIPTION
[0050] The present disclosure is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present disclosure and are not used to limit the present disclosure.
[0051] In the subsequent description, the use of suffixes such as "module", "component" or "unit" to represent elements is only to facilitate the description of the present disclosure, and has no specific meaning in itself. Therefore, "module", "component" or "unit" can be used in a mixed manner. In addition, in the subsequent description, the use of prefixes such as "first" or "second" to identify information is only to facilitate the description of the present disclosure, and has no specific meaning in itself. In addition, in the subsequent description, "at least one" means one or more; "multiple" means two or more. "At least one" is one or more; "multiple" means two or more.
[0052] At present, the recommendation system plays a huge role in consumption scenarios such as information flow and commodity flow. It records the user's explicit or implicit consumption behavior and depicts the user's portrait. Combined with the user portrait and algorithm model, it finally recommends the user's suitable materials for exposure and consumption through layer-by-layer screening from a massive candidate set. Before a material is exposed on the user side, it will go through multiple processes (or steps) such as screening and warehousing, multi-channel recall, rough sorting, fine sorting, re-sorting and / or strategy adjustment. In each process, the model and strategy may coexist and influence each other, and overlap with each other, resulting in a complex overall process and a long link.
[0053] In the current recommendation system, there are some common demand scenarios, such as online problem location, new function experience and / or recommendation result traceability, which require intuitive data display or construction of specific environments for the full-link functions of the recommendation system. For example, for online problem location or fault analysis, currently, it is often through online monitoring, local construction of test service instances and other methods to try to reproduce the problem afterwards and locate the cause of the problem. For example, for the addition of new strategies and new functions, in the testing and experience phase before going online, it is necessary to link the upstream and downstream link modules, construct material input, user's hardware and software environment, historical behavior, interest tags and other data to trigger the real effect of the new function. Complex requirements may require developers to temporarily develop relevant logic in the code to meet the conditions of the new features in order to experience the final effect. For example, for the traceability of recommendation results, it is also necessary to make corresponding engineering modifications from the most downstream module to the entire link of the user client, and support the output display of link data for a single request, so as to intuitively see the screening process of the recommendation results at each step.
[0054] At present, in order to meet the demand for tracing the source of recommendation results, each module needs to develop specific code logic, such as printing related logs, adding monitoring and other operations. For the demand for new strategies or new functions, it is necessary to find special users who meet the conditions to make requests; for example, there is a product strategy that hopes to recommend local hot spots or food attractions to users in off-site travel scenarios (travel, business trips, etc.). Only people whose permanent residence characteristics have changed meet the policy conditions. If the user experience of the policy is specified, the user's permanent residence characteristics need to be modified before requesting; similar modifications will involve many user characteristics, such as network environment, software and hardware environment, request source, application (Application, App) version, specific interests and hobbies, historical behavior, etc., resulting in the need to develop specific code, test, and deploy services online every time there is a new demand, which consumes a lot of resources and has low operating efficiency.
[0055] Based on this, the present disclosure provides a material recommendation method. Figure 1 A flow chart of a material recommendation method provided in an embodiment of the present disclosure, such as Figure 1 As shown, the following steps are included:
[0056] S11, acquiring feature data configured for a target user; wherein the feature data includes user features and / or material features;
[0057] S12: In response to detecting a recommendation request for the target user, determining a user profile of the target user and target material features of the preset material library based on the configured feature data and the initial material features of the preset material library; wherein the configured user features are used to construct the user profile of the target user, and the configured material features are used to update the initial material features of the preset material library to generate the target material features;
[0058] S13: Perform recommendation processing on the recommendation request based on the user portrait of the target user and the characteristics of the target material to obtain a target material recommended for the target user.
[0059] In the embodiment of the present disclosure, the material recommendation method can be executed by an electronic device, and the electronic device can be a server or a terminal device, which is not limited by the embodiment of the present disclosure.
[0060] In step S11, the electronic device obtains characteristic data configured for the target user, and the target user may be a simulated user in an offline simulated recommendation environment or an online user in an online recommendation environment. The target user may be a user of a social application, a user of a shopping application, a user of a fitness application or a game application, etc., which is not limited in the embodiments of the present disclosure.
[0061] In the disclosed embodiments, the electronic device may receive feature data configured for the target user through voice commands, or may receive feature data configured for the target user based on a displayed configuration page, or the electronic device may receive feature data sent by other devices.
[0062] In the disclosed embodiments, feature data includes user features and / or material features, wherein user features are features used to describe various attributes and behaviors of users, and user features may include natural attribute features of users, such as name, gender, age, region and other basic information; user features may also include user interests, crowd attributes (such as students, office workers, etc.), social relationships, etc.; in addition, user features may also include user registration, activity, payment, browsing, clicking, purchase and other behavioral records on the platform, as well as the user's preference for items, etc. User features are an important data source and basis for building user portraits, which can be understood as a comprehensive analysis and visual expression of user features.
[0063] In the disclosed embodiments, material features are features used to describe the properties and characteristics of materials. Material features may include basic properties of materials, such as classification (pictures, texts or videos, etc.), content themes, prices and other basic information; material features may also include properties that change over time, such as material popularity and topic popularity. These features are of great significance for capturing the popular trends of materials and the dynamic changes of user interests; in addition, material features may also include similarity with other materials, such as topic similarity, classification matching or associated search keywords, etc.
[0064] In the recommendation system, user characteristics and material characteristics are important foundations for building recommendation methods. These characteristics help the system understand the user's interests and behavior patterns, as well as the properties and characteristics of materials, so as to more accurately recommend relevant materials to users. Taking model-based recommendation as an example, the user characteristics for building a recommendation model can be the user characteristics of many sample users, and the corresponding material characteristics can be the characteristics of materials that the sample users have or have not had behavior records. Machine learning methods can be used to mine the relationship between user characteristics and material characteristics, thereby building a model so that materials that may be of interest to users can be recommended during the use of the model.
[0065] In the embodiment of the present disclosure, the user features configured for the target user are features associated with the target user, and the user features can be used to construct a user profile of the target user; the material features configured for the target user can indicate, for example, what type of material has what kind of popularity, etc. Exemplarily, the configured feature data can adopt a json structure, where the key is the material attribute name or the user feature keyword, and the value is a new material feature or a new user feature.
[0066] Figure 2 This is an example diagram of a page configuration feature data in an embodiment of the present disclosure, such as Figure 2 As shown, the part shown in 21 is the feature data in the configured json format, where "self_gender:fm" is the user feature keyword, and "fm" is the user feature. This configuration indicates that the target user is female.
[0067] In step S12, the electronic device responds to the detection of the recommendation request for the target user, that is, based on the configured feature data and the initial material features of the preset material library, determines the user portrait of the target user and the target material features of the preset material library. It should be noted that the initial material features are the features of each material in the preset material library, the preset material library can be a collection of materials to be recommended, and the initial material features can be features extracted in advance and stored in the electronic device; the configured material features are used to update the features of materials of a certain type or theme in the preset material library, or to update the features of a specific material, etc. Among them, if there is no corresponding feature in the initial material features of the material to be updated, the configured material features can be used to supplement the initial material features; and if there is a corresponding feature in the initial material features of the material to be updated, the configured material features can be used to modify the initial material features. In the embodiment of the present disclosure, the initial material features of the preset material library can be updated by the configured material features, so as to obtain the target material features of the preset material library.
[0068] In step S13, the electronic device performs recommendation processing based on the user portrait of the target user and the characteristics of the target material to obtain a target material recommended for the target user. The target material is one or more materials suitable for the target user selected from a preset material library.
[0069] In the disclosed embodiments, user characteristics and / or material characteristics may be configured for target users based on actual application requirements to obtain desired recommendation results or to perform relevant analysis based on the recommendation results.
[0070] For example, for out-of-town travel scenarios (tourism, business trips, etc.), a new function is desired to recommend local hot spots or food attractions to users. Only people whose permanent residence characteristics have changed meet the policy conditions. In the embodiments of the present disclosure, the user characteristics can be configured to include the changed address without modifying the user's address characteristics in the original code.
[0071] For example, for some user characteristics that are easy to change, including the network environment, software and hardware environment of the device used by the user, the application (Application, App) version used by the user, new interests and hobbies, etc., in the embodiments of the present disclosure, a new user portrait can be constructed based on the configuration of the target user's current latest user characteristics without modifying the original code.
[0072] For another example, the characteristics of the materials in the preset material library may also change at different times. For example, during the Olympic Games, the popularity of sports-related materials will become higher. Then, based on the method of the embodiment of the present disclosure, the popularity of sports-related materials can be configured to be the highest, so that the probability of sports-related materials being recommended is increased when recommending materials. Through the method of the embodiment of the present disclosure, material recommendations can be easily adapted to popular trends without modifying the original code.
[0073] In addition, the solution of the embodiment of the present disclosure is not only applicable to actual online recommendations, but also to the retrospective needs of recommendations to determine the accuracy of recommendations. For example, for any actual user, the user characteristics of the user (target user) can be simulated and configured to obtain the target material recommended for the user, so as to facilitate checking whether the recommended target material is suitable for the user. For another example, the embodiment of the present disclosure supports setting the target user as a user with a sports hobby, and setting the popularity of sports materials to be the highest. Then, after obtaining the target material based on the solution of the embodiment of the present disclosure, check whether the target material recommended for the target user is indeed a sports material. Through the method of the embodiment of the present disclosure, there is no need to develop specific code logic for simulated recommendations.
[0074] It can be understood that the embodiments of the present disclosure support the configuration of user features and / or material features. Since both user features and material features are the features most directly related to the recommendation results, the embodiments of the present disclosure optimize the problem from the underlying data dimension, making it convenient to simulate any recommendation requirements (such as scene reproduction or new function experience) through configuration without redeveloping the recommendation logic. This can reduce the problem of high resource usage and low operating efficiency caused by redeveloping the original recommendation logic.
[0075] In some embodiments, in response to detecting a recommendation request for the target user, determining a user profile of the target user and target material characteristics of the preset material library based on the configured feature data and initial material characteristics of the preset material library includes:
[0076] In response to detecting a recommendation request for the target user in an offline simulated recommendation environment, and the configured feature data includes the user features and the material features, using the configured user features to replace the actual user features of the target user and constructing a user profile of the target user, and updating the configured material features to the initial material features to obtain the target material features;
[0077] In response to detecting a recommendation request for the target user in an offline simulated recommendation environment, and the configured feature data includes the user features, the configured user features are used to replace the actual user features of the target user and a user profile of the target user is constructed, and the initial material features are used as the target material features.
[0078] In the embodiment of the present disclosure, in the offline simulated recommendation environment, since it is not an actual recommendation requirement, it is necessary to configure user features to build a simulated user profile of the target user, and the preset material library has its corresponding initial material features, so the configuration of material features is not necessary and can be determined according to the actual recommendation requirements. In the embodiment of the present disclosure, if the material features are configured, the initial material features of the material are supplemented or modified according to the material features of the configured material, so as to obtain the target material features of the preset material library; if the material features are not configured, the initial material features are directly used as the target material features.
[0079] It can be understood that in the embodiments of the present disclosure, the content of the configured feature data is processed in a targeted manner in an offline model environment so as to adapt to various recommendation requirements and the solution is highly intelligent.
[0080] In some embodiments, in response to detecting a recommendation request for the target user, determining a user profile of the target user and target material characteristics of the preset material library based on the configured feature data and initial material characteristics of the preset material library includes:
[0081] In response to detecting a recommendation request for the target user in an online recommendation environment, and the configured feature data includes the user features and the material features, constructing a user profile of the target user based on original user features of the target user obtained online and the configured user features, and updating the configured material features to the initial material features to obtain the target material features;
[0082] In response to detecting a recommendation request for the target user in an online recommendation environment, and the configured feature data includes the user features, constructing a user profile of the target user based on original user features of the target user obtained online and the configured user features, and using the initial material features as the target material features;
[0083] In response to detecting a recommendation request for the target user in an online recommendation environment, and the configured feature data includes the material features, a user profile of the target user is constructed based on the original user features of the target user obtained online, and the configured material features are updated to the initial material features to obtain the target material features.
[0084] In the embodiment of the present disclosure, in the online recommendation environment, user features and / or material features may also be configured to make the recommendation more accurate. In the embodiment of the present disclosure, since the electronic device in the online recommendation environment can obtain the user name of the online target user, the electronic device can obtain the original user features of the target user based on the user name, for example, the original user features can be determined based on the target user registration information, personal signature or historical behavior records. If the configured feature data includes user features, the electronic device can supplement or modify the original user features according to the configured user features, thereby constructing a user portrait of the target user based on the updated user features. Exemplarily, the configured user features may indicate a change in the user's permanent place of residence, or indicate a new hobby or interest of the user. If the configured feature data does not include user features, the electronic device can construct a user portrait based on the original user features, and can also implement material recommendations for the target user.
[0085] In the embodiment of the present disclosure, in the offline simulation recommendation environment, if material characteristics are configured, the initial material characteristics of the material are supplemented or modified according to the material characteristics of the configured material, so as to obtain the target material characteristics of the preset material library; if material characteristics are not configured, the initial material characteristics are directly used as the target material characteristics.
[0086] It can be understood that in the embodiments of the present disclosure, in the online recommendation environment, the content of the configured feature data is processed in a targeted manner so as to adapt to various recommendation requirements, and the solution is highly intelligent.
[0087] In some embodiments, the determining the user portrait of the target user and the target material characteristics of the preset material library based on the configured feature data and the initial material characteristics of the preset material library includes:
[0088] The configured feature data is adapted based on the recommended processing requirements to obtain adapted feature data; the adapted feature data includes adapted user features and / or material features;
[0089] Based on the adapted feature data and the initial material features of the preset material library, the user portrait of the target user and the target material features of the preset material library are determined.
[0090] In the disclosed embodiment, the configured feature number is adapted based on the recommendation processing requirements, including the data format, the measurement range of the data, etc., to adapt to the original recommendation system, that is, the configured user features and / or material features meet the data processing requirements of the original recommendation system. It should be noted that in the disclosed embodiment, in the online recommendation environment, if no feature data is configured, the electronic device can still recommend materials to the target user based on the original user features of the target user and the initial material features of the preset material library.
[0091] It can be understood that in the embodiments of the present disclosure, the adaptation process facilitates the expansion of configuration requirements without changing the original recommended code, and has good flexibility.
[0092] In some embodiments, the target user is a user with feature data configuration permission.
[0093] In the disclosed embodiment, the electronic device provides a whitelist restriction, and only whitelist users have the authority to configure feature data, wherein the whitelist users may be VIP users or paying users, etc.
[0094] It can be understood that the electronic device in the embodiment of the present disclosure allows the configuration of characteristic data for whitelist users, which can improve the accuracy of recommendations for whitelist users while saving resource usage.
[0095] In some embodiments, the method further comprises:
[0096] In the process of performing recommendation processing on the recommendation request based on the user portrait of the target user and the target material characteristics, saving the material screening results of the intermediate stage; wherein the intermediate stage includes at least one of the following: recall, rough sorting, fine sorting, and re-sorting;
[0097] The page displays the material screening results of the intermediate stage.
[0098] In the disclosed embodiment, the material recommendation method will go through the core links of recall, rough sorting, fine sorting, and re-ranking during the execution process, which work together to provide users with personalized and accurate recommendation services. Among them, recall is the first step of the recommendation system, and its main purpose is to quickly screen out items that the target user may be interested in from the preset material library; rough sorting is the transition link between recall and fine sorting, and its main purpose is to preliminarily screen and sort the results of recall, further reduce the number of candidate materials, and ensure a certain accuracy; the main task of fine sorting is to deeply score and sort the results of rough sorting, to ensure that the materials finally recommended to the user are both in line with the user's interests and have high quality; and re-ranking is the last process of the recommendation system, and its main task is to re-rank or adjust the results of fine sorting to further improve the accuracy of the recommendation and user experience. Usually, the re-ranking stage focuses on business strategies and personalized needs.
[0099] In the disclosed embodiment, it supports displaying the material screening results of multiple intermediate stages on the same page, and also supports displaying the material screening results of each intermediate stage on different pages.
[0100] It can be understood that in the embodiment of the present disclosure, by saving the material screening results of one or more intermediate stages and displaying them on the page, it is convenient to intuitively locate the problem based on the displayed material screening results.
[0101] In some embodiments, the page displays the material screening results of the intermediate stage, including:
[0102] In response to detecting an instruction to display the processing result of any intermediate stage, the page displays the material screening result of the intermediate stage indicated by the instruction.
[0103] In the disclosed embodiment, the processing results of each intermediate stage can be displayed on different pages based on the instructions for display, so as to facilitate the user to view. The instructions for display can be voice instructions or click instructions for corresponding controls.
[0104] For example, Figure 3 This is an example diagram of a page display in an embodiment of the present disclosure, such as Figure 3 As shown, the "sorting link" shown in 31 supports the selection of page display items, and can support the display of processing results of all intermediate stages, and also supports displaying only the processing results of a single intermediate stage. Figure 3 In the example, the "All" control is currently selected, and the scores of each intermediate stage are displayed in the result display part shown in 32.
[0105] It can be understood that in the embodiment of the present disclosure, the corresponding page display is performed based on instructions, which is more intelligent.
[0106] Figure 4is a schematic diagram of the material recommendation method in the embodiment of the present disclosure, such as Figure 4 As shown, the part shown at 41 is the original recommendation system, which is used to execute the process of the material recommendation algorithm; the part shown at 42 is the simulation system, which is used to configure feature data and display results. Figure 4 In the present invention, based on the whitelist on the simulation system side, the configuration personnel can select any whitelist user for simulation configuration, including configuring user characteristics and / or material characteristics; wherein, the configured user characteristics (i.e., user characteristic simulation) are used to update the behavior sequence or user attributes on the recommendation system side to update the user portrait; the configured material characteristics (i.e., material characteristic simulation) are used to update the basic material platform on the recommendation system side (i.e., the preset material library in the embodiment of the present disclosure). In the embodiment of the present disclosure, the update of the basic material platform refers to updating the initial material characteristics of the preset material library to obtain the target material characteristics.
[0107] like Figure 4 As shown, after the recommendation system side constructs the user portrait of the target user based on the configured feature data and obtains the target material features, it executes the recommendation process of multi-way recall, rough sorting, fine sorting, and rearrangement, and finally obtains the recommendation result. It should be noted that when the recommendation system side executes the above process, the master control module is responsible for reading the configured feature data and adapting it; wherein, the configured feature data can be stored in the Remote Dictionary Server (Redis) database, and the master control module can read the configured feature data from the Redis database. In addition, the recommendation system can record the behavior of the target user after the recommendation during the recommendation process, so as to update and optimize the recommendation in real time according to the behavior record results; and, the recommendation system will save the material screening results in the recall, rough sorting, fine sorting, and rearrangement process to the full-link database during the recommendation process to facilitate the subsequent page display. The data stored in the full-link database may include a list of material identifications (Identity, ID) at each stage, material tags, recall IDs, request IDs, user IDs, and ranking scores, etc., which are not limited by the embodiments of the present disclosure. Among them, the material ID list field indicates a collection of material IDs; the material tag indicates at least one tag of the material; the recall ID indicates one of the multiple recalls, and the multiple recalls can indicate the type of material, such as pictures, text or videos, etc., and different types have different corresponding IDs; the request ID is used to indicate a recommendation request; and the user ID is used to indicate a user.
[0108] Figure 5 A material recommendation device diagram is provided for an embodiment of the present disclosure, such as Figure 5 As shown, the device 500 includes:
[0109] The acquisition module 501 is configured to acquire feature data configured for a target user; wherein the feature data includes user features and / or material features;
[0110] The determination module 502 is configured to, in response to detecting a recommendation request for the target user, determine a user profile of the target user and target material features of the preset material library based on the configured feature data and the initial material features of the preset material library; wherein the configured user features are used to construct the user profile of the target user, and the configured material features are used to update the initial material features of the preset material library to generate the target material features;
[0111] The recommendation module 503 is configured to perform recommendation processing on the recommendation request based on the user portrait of the target user and the characteristics of the target material to obtain the target material recommended for the target user.
[0112] In some embodiments, the determination module 502 is further configured to: in response to detecting a recommendation request for the target user in an offline simulated recommendation environment, and the configured feature data includes the user features and the material features, use the configured user features to replace the actual user features of the target user and construct a user profile of the target user, and update the configured material features to the initial material features to obtain the target material features; in response to detecting a recommendation request for the target user in an offline simulated recommendation environment, and the configured feature data includes the user features, use the configured user features to replace the actual user features of the target user and construct a user profile of the target user, and use the initial material features as the target material features.
[0113] In some embodiments, the determination module 502 is further configured to: in response to detecting a recommendation request for the target user in an online recommendation environment, and the configured feature data includes the user features and the material features, construct a user profile of the target user based on the original user features of the target user obtained online and the configured user features, and update the configured material features to the initial material features to obtain the target material features; in response to detecting a recommendation request for the target user in an online recommendation environment, and the configured feature data includes the user features, construct a user profile of the target user based on the original user features of the target user obtained online and the configured user features, and use the initial material features as the target material features; in response to detecting a recommendation request for the target user in an online recommendation environment, and the configured feature data includes the material features, construct a user profile of the target user based on the original user features of the target user obtained online, and update the configured material features to the initial material features to obtain the target material features.
[0114] In some embodiments, the determination module 502 is further configured to: adapt the configured feature data based on the recommendation processing requirements to obtain adapted feature data; the adapted feature data includes adapted user characteristics and / or material characteristics; based on the adapted feature data and the initial material characteristics of the preset material library, determine the user portrait of the target user and the target material characteristics of the preset material library.
[0115] In some embodiments, the target user is a user with feature data configuration permission.
[0116] In some embodiments, the apparatus further comprises:
[0117] A saving module is configured to save the material screening results of the intermediate stage in the process of performing recommendation processing on the recommendation request based on the user portrait of the target user and the target material characteristics; wherein the intermediate stage includes at least one of the following: recall, rough sorting, fine sorting, and re-sorting;
[0118] The display module is configured to display the material screening results of the intermediate stage on a page.
[0119] In some embodiments, the display module is further configured to: in response to detecting an instruction to display the processing results of any intermediate stage, display the material screening results of the intermediate stage indicated by the instruction on the page.
[0120] Regarding the device in the above embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.
[0121] Figure 6 is a schematic diagram of a hardware entity of an electronic device in an embodiment of the present disclosure, such as Figure 6 As shown, the hardware entity of the computer device 600 includes: a processor 601, a communication interface 602 and a memory 603, wherein:
[0122] Processor 601 generally controls the overall operation of computer device 600 .
[0123] The communication interface 602 enables the computer device to communicate with other terminals or servers through a network.
[0124] The memory 603 is configured to store instructions and applications executable by the processor 601, and can also cache data to be processed or processed by the processor 601 and each module in the computer device 600 (for example, image data, audio data, voice communication data, and video communication data), which can be implemented by flash memory (FLASH) or random access memory (Random Access Memory, RAM). Data can be transmitted between the processor 601, the communication interface 602 and the memory 603 through the bus 604. Among them, the processor 601 is used to execute some or all of the steps in the above method.
[0125] Correspondingly, an embodiment of the present disclosure provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, part or all of the steps in the above method are implemented.
[0126] The embodiment of the present disclosure provides a computer program product, which includes: a computer program or executable instructions, which are stored in a computer-readable storage medium. The processor of the computer device reads the computer program or executable instructions from the computer-readable storage medium, and the processor executes the computer program or executable instructions, so that the computer device executes any one of the above-mentioned material recommendation methods of the embodiment of the present disclosure.
[0127] It should be noted that the description of the above storage medium and device embodiments is similar to the description of the above method embodiments, and has similar beneficial effects as the method embodiments. For technical details not disclosed in the storage medium and device embodiments of the present disclosure, please refer to the description of the method embodiments of the present disclosure for understanding.
[0128] It should be understood that in various embodiments of the present disclosure, the size of the sequence number of each process does not mean the order of execution, and 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 disclosure. The sequence numbers of the embodiments of the present disclosure are only for description and do not represent the advantages and disadvantages of the embodiments.
[0129] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or device including the element.
[0130] In the several embodiments provided in the present disclosure, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored, or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.
[0131] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units; they may be located in one place or distributed on multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.
[0132] In addition, all functional units in the embodiments of the present disclosure may be integrated into one processing unit, or each unit may be separately configured as a unit, or two or more units may be integrated into one unit; the above-mentioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.
[0133] A person skilled in the art can understand that all or part of the steps of implementing the above method embodiment can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above method embodiment; and the aforementioned storage medium includes: mobile storage devices, read-only memories (ROM), magnetic disks or optical disks, etc., various media that can store program codes.
[0134] Alternatively, if the above-mentioned integrated unit of the present disclosure is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present disclosure can essentially or in other words, the part that contributes to the relevant technology can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the methods described in each embodiment of the present disclosure. The aforementioned storage medium includes: various media that can store program codes, such as mobile storage devices, ROMs, magnetic disks, or optical disks.
[0135] The above description is only an embodiment of the present disclosure and is not intended to limit the protection scope of the present disclosure. Any modification, equivalent replacement and improvement made within the spirit and scope of the present disclosure are included in the protection scope of the present disclosure.
Claims
1. A material recommendation method, characterized in that: include: Acquire characteristic data configured for a target user; wherein the characteristic data includes user characteristics and / or material characteristics; In response to detecting a recommendation request for the target user, based on the configured feature data and the initial material features of the preset material library, determining a user profile of the target user and target material features of the preset material library; wherein the configured user features are used to construct the user profile of the target user, and the configured material features are used to update the initial material features of the preset material library to generate the target material features; The recommendation request is processed based on the user portrait of the target user and the characteristics of the target material to obtain a target material recommended for the target user.
2. The method according to claim 1, characterized in that The step of determining, in response to detecting a recommendation request for the target user, a user profile of the target user and target material characteristics of the preset material library based on the configured feature data and initial material characteristics of the preset material library, includes: In response to detecting a recommendation request for the target user in an offline simulated recommendation environment, and the configured feature data includes the user features and the material features, using the configured user features to replace the actual user features of the target user and constructing a user profile of the target user, and updating the configured material features to the initial material features to obtain the target material features; In response to detecting a recommendation request for the target user in an offline simulated recommendation environment, and the configured feature data includes the user features, the configured user features are used to replace the actual user features of the target user and a user profile of the target user is constructed, and the initial material features are used as the target material features.
3. The method according to claim 1, characterized in that The step of determining, in response to detecting a recommendation request for the target user, a user profile of the target user and target material characteristics of the preset material library based on the configured feature data and initial material characteristics of the preset material library, includes: In response to detecting a recommendation request for the target user in an online recommendation environment, and the configured feature data includes the user features and the material features, constructing a user profile of the target user based on original user features of the target user obtained online and the configured user features, and updating the configured material features to the initial material features to obtain the target material features; In response to detecting a recommendation request for the target user in an online recommendation environment, and the configured feature data includes the user features, constructing a user profile of the target user based on original user features of the target user obtained online and the configured user features, and using the initial material features as the target material features; In response to detecting a recommendation request for the target user in an online recommendation environment, and the configured feature data includes the material features, a user profile of the target user is constructed based on the original user features of the target user obtained online, and the configured material features are updated to the initial material features to obtain the target material features.
4. The method according to any one of claims 1 to 3, characterized in that The determining, based on the configured characteristic data and the initial material characteristics of the preset material library, the user portrait of the target user and the target material characteristics of the preset material library comprises: The configured feature data is adapted based on the recommended processing requirements to obtain adapted feature data; the adapted feature data includes adapted user features and / or material features; Based on the adapted feature data and the initial material features of the preset material library, the user portrait of the target user and the target material features of the preset material library are determined.
5. The method according to claim 1, characterized in that The target user is a user with feature data configuration authority.
6. The method according to claim 1, characterized in that The method further comprises: In the process of performing recommendation processing on the recommendation request based on the user portrait of the target user and the target material characteristics, saving the material screening results of the intermediate stage; wherein the intermediate stage includes at least one of the following: recall, rough sorting, fine sorting, and re-sorting; The page displays the material screening results of the intermediate stage.
7. The method according to claim 6, characterized in that The page displays the material screening results of the intermediate stage, including: In response to detecting an instruction to display the processing result of any intermediate stage, the page displays the material screening result of the intermediate stage indicated by the instruction.
8. A material recommendation device, characterized in that: include: An acquisition module, configured to acquire feature data configured for a target user; wherein the feature data includes user features and / or material features; a determination module configured to, in response to detecting a recommendation request for the target user, determine a user profile of the target user and target material features of the preset material library based on the configured feature data and the initial material features of the preset material library; wherein the configured user features are used to construct the user profile of the target user, and the configured material features are used to update the initial material features of the preset material library to generate the target material features; The recommendation module is configured to perform recommendation processing on the recommendation request based on the user portrait of the target user and the characteristics of the target material to obtain the target material recommended for the target user.
9. An electronic device, characterized in that: The electronic device comprises: A memory for storing executable instructions; A processor, configured to implement the material recommendation method according to any one of claims 1 to 7 when executing the executable instructions stored in the memory.
10. A computer storage medium, characterized in that: The computer storage instructions store a computer program or an executable instruction, wherein the computer program or the executable instruction, when executed by a processor, implements the material recommendation method according to any one of claims 1 to 7.
