Service recommendation method and device, electronic equipment and computer readable storage medium

Abstract

The embodiment of the invention discloses a service recommendation method and device, electronic equipment and a computer readable storage medium, and belongs to the technical field of recommendation. The method comprises: in response to a current key event for a target device, determining service preference information of a target user, and determining use attribute information of the target device, the use attribute information comprising scene information and time period information; and outputting target service recommendation information based on the service preference information and the use attribute information. Thus, when the key operation of the target user is detected, the user service preference and the equipment use attribute can be automatically recognized, intelligent and accurate personalized service recommendation is achieved, the function boundary of physical keys is expanded, the intelligent level of service ecology is effectively improved, and then the interaction efficiency and the user experience are remarkably improved.

Description

Service recommendation methods, devices, electronic devices and computer-readable storage media Technical Field

[0001] This application relates to the field of recommendation technology, specifically to a service recommendation method, apparatus, electronic device, and computer-readable storage medium. Background Technology

[0002] With the widespread adoption of smart TVs, set-top boxes, and other terminals, physical buttons serve as a stable interaction point, primarily used for basic control operations such as channel switching and volume adjustment.

[0003] However, current physical buttons have limited functionality, only able to trigger preset commands and unable to provide personalized service recommendations, thus restricting the improvement of user experience. Summary of the Invention

[0004] This application provides a service recommendation method, apparatus, electronic device, and computer-readable storage medium that can respond to current key events to achieve intelligent and accurate personalized service recommendations, significantly enhancing interaction efficiency and user experience.

[0005] In a first aspect, embodiments of this application provide a service recommendation method, comprising: in response to a current key event on a target device, determining service preference information of a target user and determining usage attribute information of the target device; wherein the usage attribute information includes scenario information and time period information; and outputting target service recommendation information based on the service preference information and the usage attribute information.

[0006] In one embodiment, the current key event includes key type information; determining the target user's service preference information includes: acquiring the target user's user characteristic information; and determining the target user's service preference information based on the user characteristic information and the key type information.

[0007] In one embodiment, determining the usage attribute information of the target device includes: obtaining the currently playing content of the target device; identifying the currently playing content to obtain the scene information and the time period information; wherein the scene information represents the content scene information of the currently playing content, and the time period information represents the content time period information of the currently playing content.

[0008] In one embodiment, the target service recommendation information includes multiple recommended service contents; the step of outputting the target service recommendation information based on the service preference information and the usage attribute information includes: determining multiple candidate service contents based on the service preference information and the usage attribute information; obtaining the target user's operation preference information for the multiple candidate service contents; determining multiple recommended service contents and their recommendation priorities among the multiple candidate service contents based on the operation preference information and the scenario information; and outputting the multiple recommended service contents based on the recommendation priorities.

[0009] In one embodiment, after outputting multiple recommended service contents based on the recommendation priority, the method further includes: in response to an operation instruction for any of the multiple recommended service contents, executing the functional service corresponding to that recommended service content.

[0010] In one embodiment, determining multiple candidate service contents based on the service preference information and the usage attribute information includes: inputting the service preference information and the usage attribute information into a preset processing model to obtain multiple candidate service contents output by the processing model.

[0011] In one embodiment, after outputting multiple recommended service contents based on the recommendation priority, the method further includes: recording operation result information of the multiple recommended service contents; updating operation preference information of the multiple recommended service contents based on the operation result information; and / or uploading the operation result information to the cloud and updating the model parameters of the processing model based on the model update parameters returned by the cloud; wherein the model update parameters are determined by the cloud to train a cloud model based on the operation result information and based on the updated model parameters of the cloud model.

[0012] Secondly, embodiments of this application provide a service recommendation device, the device comprising: an information determination module, configured to determine service preference information of a target user and usage attribute information of the target device in response to a current key event of a target device; wherein the usage attribute information includes scenario information and time period information; and a service recommendation module, configured to output target service recommendation information based on the service preference information and the usage attribute information.

[0013] In one embodiment, the current key event includes key type information; the information determination module includes: a user information acquisition submodule, used to acquire user feature information of the target user; and a service preference determination submodule, used to determine the service preference information of the target user based on the user feature information and the key type information.

[0014] In one embodiment, the information determination module includes: a content acquisition submodule, used to acquire the currently playing content of the target device; and a content recognition submodule, used to recognize the currently playing content to obtain the scene information and the time period information; wherein the scene information represents the content scene information of the currently playing content, and the time period information represents the content time period information of the currently playing content.

[0015] In one embodiment, the target service recommendation information includes multiple recommended service contents; the service recommendation module includes: a candidate content determination submodule, used to determine multiple candidate service contents based on the service preference information and the usage attribute information; a preference information acquisition submodule, used to acquire the target user's operation preference information for the multiple candidate service contents; a recommendation content determination submodule, used to determine multiple recommended service contents and their recommendation priorities among the multiple candidate service contents based on the operation preference information and the scenario information; and a recommendation content output submodule, used to output multiple recommended service contents based on the recommendation priorities.

[0016] In one embodiment, the service recommendation device further includes a service execution module, configured to execute the functional service corresponding to any one of the multiple recommended service contents in response to an operation instruction for any one of the recommended service contents.

[0017] In one embodiment, the candidate content determination submodule is specifically used to input the service preference information and the usage attribute information into a preset processing model to obtain a plurality of candidate service contents output by the processing model.

[0018] In one embodiment, the service recommendation device further includes: an operation recording module for recording operation result information of multiple recommended service contents; a preference update module for updating operation preference information of multiple recommended service contents based on the operation result information; and a model update module for uploading the operation result information to the cloud and updating the model parameters of the processing model based on the model update parameters returned by the cloud; wherein the model update parameters are determined by the cloud based on the operation result information, training a cloud model, and the updated model parameters of the cloud model.

[0019] Thirdly, embodiments of this application also provide an electronic device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, it implements the steps in the above-described service recommendation method.

[0020] Fourthly, embodiments of this application also provide a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps in the service recommendation method described above.

[0021] Fifthly, embodiments of this application also provide a computer program product or computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the methods provided in the various optional implementations described in the embodiments of this application.

[0022] In summary, in this embodiment, in response to a current key press event on the target device, the service preference information of the target user and the usage attribute information of the target device can be determined. The usage attribute information includes scenario information and time period information. Based on the service preference information and usage attribute information, target service recommendation information is output. Thus, when a key press operation of the target user is detected, user service preferences and device usage attributes can be automatically identified, achieving intelligent and accurate personalized service recommendations. This expands the functional boundaries of physical buttons while effectively improving the intelligence level of the service ecosystem, thereby significantly enhancing interaction efficiency and user experience. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 is a flowchart illustrating a service recommendation method according to an embodiment of this application; Figure 2 is a flowchart illustrating a specific embodiment of determining service preference information according to an embodiment of this application; Figure 3 is a flowchart illustrating a specific embodiment of outputting target service recommendation information according to an embodiment of this application; Figure 4 is a schematic diagram illustrating a display interface of multiple recommended service contents according to an embodiment of this application; Figure 5 is a flowchart illustrating a specific embodiment of updating operation preference information and processing model according to an embodiment of this application; Figure 6 is a schematic diagram illustrating the structure of a service recommendation device according to an embodiment of this application; Figure 7 is a schematic diagram illustrating the structure of an electronic device according to an embodiment of this application. Detailed Implementation

[0025] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0026] It's important to note that in the field of smart terminals, physical buttons (such as remote control buttons) are widely used as a traditional interaction method for performing basic function controls, such as channel switching, volume adjustment, or menu access. However, existing technologies have significant limitations: First, the functions of physical buttons are fixed, only able to trigger preset commands, and cannot provide adaptive service recommendations based on the content playback progress (such as the intro, main feature, and outro of a movie or TV show) or real-time user behavior. For example, some solutions involve scene-based content recommendations, but fail to dynamically link physical button operations with content time periods, resulting in a passive and untargeted recommendation mechanism.

[0027] Secondly, existing service recommendation systems largely rely on cloud-based big data analytics for automated push notifications (such as advertisements and membership reminders). However, these notifications are mostly in the form of single cards or static menus, unable to respond to physical button events to achieve batch, multi-service access linkage. For example, while some solutions attempt to recommend content based on user behavior data, their push mechanisms are simplistic and fail to integrate content time-based tags with physical button triggering logic, resulting in a disconnect between the recommendation service and users' real-time needs.

[0028] In addition, although some smart TVs currently have content awareness capabilities, their recommendation services are often limited to platform algorithm-driven approaches and fail to effectively utilize physical buttons as active interaction entry points, resulting in cumbersome operation, fragmented service entry points, and poor user experience.

[0029] Overall, existing technologies struggle to intelligently integrate physical buttons with content time periods and user behavior, limiting the flexibility of smart terminal interaction and the accuracy of service recommendations.

[0030] To address the issue that current physical buttons offer limited functionality and cannot provide personalized service recommendations, this application aims to offer a service recommendation method. In response to a current key press event on a target device, the method determines the target user's service preferences and the target device's usage attributes, including scenario and time period information. Based on these service preferences and usage attributes, it outputs target service recommendations. This approach automatically identifies user service preferences and device usage attributes upon detecting a key press, enabling intelligent and accurate personalized service recommendations. While expanding the functional boundaries of physical buttons, it effectively enhances the intelligence level of the service ecosystem, thereby significantly improving interaction efficiency and user experience.

[0031] The following sections provide detailed descriptions of each example. It should be noted that the order in which the embodiments are described is not intended to limit the priority of the embodiments.

[0032] Figure 1 schematically illustrates a flowchart of a service recommendation method according to an embodiment of this application. The execution entity of this service recommendation method can be a service recommendation device, which can be integrated into any electronic device with data processing, network communication, and program execution functions. The electronic device can be a server or a terminal, etc.

[0033] The server can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, network acceleration services (Content Delivery Network, CDN), as well as big data and artificial intelligence platforms.

[0034] The terminal can be a smart TV, smart speaker, smartphone, tablet, laptop, desktop computer, etc., but is not limited to these. The terminal and the server can be connected directly or indirectly through wired or wireless communication, which is not limited herein.

[0035] Furthermore, in the embodiments of this application, "multiple" refers to two or more. The terms "first" and "second," etc., in the embodiments of this application are used for distinguishing descriptions and should not be construed as implying relative importance.

[0036] The following sections provide detailed descriptions of each example. It should be noted that the order in which the embodiments are described is not intended to limit the preferred order of the embodiments.

[0037] In this embodiment, the description will be from the perspective of a service recommendation device, which can be integrated into a server or terminal. To facilitate the explanation of the service recommendation method of this application, the following will describe the service recommendation device integrated into a terminal in detail, that is, the terminal will be used as the execution subject for detailed explanation.

[0038] Referring to Figure 1, a flowchart of a service recommendation method according to this application is shown. The method may specifically include steps S101 to S102, as follows: S101: In response to the current key event of the target device, determine the service preference information of the target user and determine the usage attribute information of the target device.

[0039] In this embodiment, the target device is a device that interacts with the user, such as a smart TV, smart speaker, wearable device, smart access control system, home control system, in-vehicle infotainment system, etc., which provides various functional services to the user.

[0040] In this embodiment, the current button event represents a button operation triggered by a target user on a target button of a target device. The target button can be a physical button or a virtual button provided by the user interface. For example, the target button can be a physical button on a smart TV remote control, a physical button on a smart speaker, a control button for each smart home device in a home control system, or a virtual button on the user interface of a wearable device. The button operation can specifically include single click, long press, double click, and combination key operations.

[0041] In this embodiment, service preference information is used to characterize the preference information for various functional services provided by the target device. This service preference information reflects the target user's degree of liking for various functional services.

[0042] In this embodiment, the target user's service preference information can be determined based on the target user's historical behavior data. This historical behavior data may include service usage records, on-demand records, favorites lists, search keywords, and other data. For example, analysis may reveal that the target user has recently been frequently watching martial arts films and television dramas and has previously saved short videos analyzing weapons. Therefore, it can be determined that the target user's service preference information indicates a stronger preference for "martial arts themes" and "weapon knowledge" when watching films and television dramas.

[0043] In this embodiment, attribute information is used to characterize the attribute information of the target device during the target user's use of the target device. This usage attribute information may specifically include scene information and time period information.

[0044] In this embodiment, scene information refers to the scene information when a target user uses the target device. For example, when the target device is a smart TV, the scene information can represent a movie scene when the user plays a movie on the smart TV; when the target device is a smart speaker, the scene information can represent a music scene when the user uses the music playback function of the smart speaker; when the target device is an in-vehicle infotainment system, the scene information can represent a car radio scene when the user uses the radio playback function of the in-vehicle infotainment system; and when the target device is an in-vehicle navigation system, the scene information can represent an in-vehicle navigation scene when the user uses the navigation function of the in-vehicle navigation system.

[0045] In this embodiment, the time period information represents the time period information when the target user uses the target device. Specifically, the time period information can represent different time periods, such as morning, noon, evening, family entertainment time, rest time, etc.; the time period information can also represent different usage stages, for example, in the case of in-vehicle navigation, the time period information can represent before driving, during driving, and after driving, etc.; the time period information can also represent the content time period information of the currently playing content on the target device, such as the beginning, middle, and end of a movie.

[0046] In this embodiment, the service recommendation function is triggered by the user's current key press event. That is, the service recommendation function is triggered in response to user operation, rather than being forcibly pushed by the system, thus greatly enhancing the initiative and controllability of the user experience. For example, the current key press event could be a user long-pressing the menu button on a smart TV remote. The smart TV detects this current key press event and provides relevant service recommendations to the user.

[0047] In this embodiment, when a current key event is detected, the terminal will identify the current service preference information, scenario information, and time period information from three dimensions: user preference, scenario, and time period, thereby providing comprehensive and accurate data support for subsequent service recommendations.

[0048] S102: Based on service preference information and usage attribute information, output target service recommendation information.

[0049] In this embodiment, service preference information and usage attribute information can be fused and calculated, and based on the calculation results, the most relevant target service recommendation information can be selected from the service resource library. The target service recommendation information may include multiple recommended service items, which can be presented in the form of service cards.

[0050] In one example, when a target user is watching a TV series on a smart TV and presses and holds the "menu" button on the remote control during playback, the system triggers intelligent service recommendations. First, the system immediately determines the target user's service preferences and identifies the content of the currently playing TV series to determine scene information (e.g., a period drama fight scene) and time period information (e.g., the main feature). Then, the system integrates this information and outputs a set of precisely matched target service recommendations on the TV screen, such as recommending multiple service cards like "similar martial arts dramas," "introduction to the martial arts choreography of this series," or "official merchandise."

[0051] In another example, when a target user is exercising with a smartwatch and long-presses the "menu" button on the smartwatch during the exercise, the system triggers intelligent service recommendations. First, the system immediately determines the target user's service preferences, while simultaneously detecting scene information (e.g., jogging) and time information (e.g., warming up) through motion sensors. Then, the system integrates this information and outputs a set of precisely matched target service recommendations on the smartwatch screen, such as multiple service cards like "heart rate monitoring," "pace monitoring," "exercise trajectory," "quick reply," and "emergency call."

[0052] In this implementation, by deeply integrating the button—a stable interaction entry point—with service preferences, scenarios, and time periods, a service recommendation mechanism that shifts from "passive response" to "proactive intelligence" is achieved. Users only need a single simple button press during device use to obtain a series of service recommendations highly relevant to the current scenario and their own interests. This greatly simplifies the operation path, enriches the interactive content, expands the functional boundaries of physical buttons, effectively improves the intelligence level of the service ecosystem, and thus significantly enhances interaction efficiency and user experience.

[0053] In one feasible implementation, the current key event includes key type information. Referring to Figure 2, the step of determining the target user's service preference information may specifically include sub-steps S201-S202, as follows: S201: Obtain the target user's user characteristic information; in this implementation, the user characteristic information may include user biometric information and / or behavioral characteristic information. Biometric information may include age, gender, etc.; behavioral characteristic information is used to characterize the target user's behavioral preferences, and this behavioral characteristic information can be obtained based on the target user's historical behavioral data analysis. For example, user characteristic information can be represented as multi-dimensional feature tags. For example, user characteristic information can be represented as: {Age: 25; Gender: Male; Behavioral characteristics: Primary preference: watching science fiction movies; Secondary preference: collecting movie soundtracks}.

[0054] In this embodiment, the target user can be a user preset by the system. For example, the target user can represent a user bound to the currently logged-in account. The target user can also be a user obtained by the terminal through image recognition technology to identify the user's image, or a user obtained by the terminal through voiceprint recognition technology to identify the user's voice.

[0055] S202: Determine the target user's service preference information based on user characteristic information and key type information.

[0056] In this embodiment, button type information is used to characterize the service intent of the target user. That is, different button type information can correspond to different service intents. This button type information can be determined based on the function type and operation method of the target button. Among them, the function type includes OK, menu, direction, etc., and the operation method includes short press (single click), long press, double click, combination key, etc.

[0057] In this embodiment, each operation mode of the target button is assigned a specific type identifier. For example, when the user briefly presses the "OK" button on the remote control, the button type of the current button event can be determined as "short press - OK button". When the user presses and holds the "OK" button on the remote control for more than 1 second, the button type of the current button event can be determined as "long press - OK button".

[0058] In this embodiment, the terminal can pre-configure a first service preference mapping table to represent the correspondence between user characteristic information, key type information, and service preference information. After obtaining the user characteristic information and key type information, the service preference information of the target user can be determined by querying the first service preference mapping table.

[0059] In this embodiment, the terminal may also pre-configure a second service preference mapping table to represent the correspondence between key type information and initial service preference information. After obtaining the key type information, the initial service preference information of the target user can be determined by querying the second service preference mapping table; then, based on the user feature information and the initial service preference information, the target user's service preference information can be determined.

[0060] In one example, when a user briefly presses the "OK" button on the remote while watching a movie, the button type information "short press - OK button" is mapped to the corresponding initial service preference information according to the second service preference mapping table. This initial service preference information is: "Trigger functional services related to the video content." Then, the user's characteristic information is checked as {Age: 25; Gender: Male; Behavioral characteristics: Primary preference: watching science fiction movies; Secondary preference: collecting movie soundtracks}. This user characteristic information is combined with the initial service preference information, and the final determined service preference information is: {Strong preference: director information, science fiction background analysis; Weak preference: collecting movie soundtracks}. That is, it is determined that the current user wants to obtain knowledge-based services that are deeply related to this science fiction movie itself.

[0061] In this embodiment, by combining button type information with user characteristic information, accurate identification of service preference information can be achieved. This allows users to express different service intentions at different times through different button gestures (such as short presses and long presses), and the terminal can accurately understand the user's service intentions and filter out the most relevant part from the user's rich characteristic profile, thereby generating service preference information that meets the user's personalized needs, effectively improving the efficiency of human-computer interaction and the satisfaction of the intelligent experience.

[0062] In one feasible implementation, the step of determining the usage attribute information of the target device may specifically include: obtaining the current playback content of the target device; identifying the current playback content to obtain scene information and time period information.

[0063] In this embodiment, the currently playing content refers to the multimedia content being played on the target device, which can be a video and / or audio stream. Specifically, the currently playing content of the target device can be obtained in real time by accessing the graphics card buffer, the decoding output of the media player, or by using screenshot technology.

[0064] In this embodiment, computational vision and audio processing technologies can be used to identify the currently playing content and obtain scene information and time period information. Scene information refers to the content scene information of the currently playing content; time period information refers to the content time period information of the currently playing content. The content scene information can specifically include the content type and identified page elements. The content type can include TV dramas, variety shows, movies, sports events, etc., and the page elements can include the characters, names, directors, and other related figures or works of the currently playing audio-visual content.

[0065] Specifically, for scene information, video frames of the currently playing content can be extracted and then input into a pre-trained image recognition model to obtain scene information output by the model. For time period information, the content structure, audio, and / or playback progress information of the currently playing content can be analyzed to obtain time period information. For example, based on the comparison of content structure with program templates, key segments such as opening / closing animations, commercial breaks, and chapter transitions can be identified; changes in background music can be identified based on audio to determine whether there are ending sound effects, the intensity of audience cheers, etc.; and time period information can be determined based on the ratio between playback progress information and the total duration of the program.

[0066] In this embodiment, by identifying the currently playing content in real time, scene information and time period information can be dynamically identified, realizing automated, high-precision, and dynamic perception of the target device's usage attribute information. This allows subsequent service recommendations to be deeply integrated with the user's real-time viewing experience, enhancing the real-time nature and accuracy of recommendations and further improving the user experience.

[0067] In one feasible implementation, the target service recommendation information includes multiple recommended service contents. Referring to Figure 3, the step of outputting the target service recommendation information based on service preference information and usage attribute information may specifically include sub-steps S301 to S304, as follows: S301: Determine multiple candidate service contents based on service preference information and usage attribute information.

[0068] In this embodiment, after determining service preference information and usage attribute information, the terminal can first perform preliminary matching and retrieval in a large general service database based on the service preference information and usage attribute information, generating a wide range of candidate service contents that may be related to the user's needs. The general service database contains various functional services that the target device can provide.

[0069] In this embodiment, multiple candidate service contents can be determined based on rules. Specifically, the step of determining multiple candidate service contents based on service preference information and usage attribute information may include: determining multiple candidate service contents based on the preset correspondence between service preference information and usage attribute information and candidate service contents.

[0070] In this embodiment, multiple candidate service contents can also be determined through model recognition. Specifically, the step of determining multiple candidate service contents based on service preference information and usage attribute information may include: inputting the service preference information and usage attribute information into a preset processing model to obtain multiple candidate service contents output by the processing model.

[0071] In this embodiment, the processing model refers to a model pre-trained on the service recommendation task. This processing model is a mathematical model capable of learning complex patterns and relationships from data; for example, it can employ a machine learning model or a deep learning model.

[0072] In this embodiment, before formally deploying the processing model to the terminal for service recommendation, the processing model can be trained on an initial model based on multiple sample data. Each sample data includes sample service preference information and sample usage attribute information, and the sample data carries service content tags. By training the initial model with a large amount of sample data, the trained processing model can learn the complex patterns between service preference information, usage attribute information, and candidate service content, thereby predicting multiple candidate service contents that the user is most likely to be interested in.

[0073] In this embodiment, feature extraction can be performed on service preference information and usage attribute information to obtain a feature vector. This feature vector can be a multi-dimensional vector encoded by the service preference information and usage attribute information. For example, the feature vector can contain vectors with hundreds of dimensions, each dimension representing a preference label or scene label. The feature vector is then input into a processing model to obtain a sorted list or a set of probability scores output by the processing model, corresponding to the IDs or vectors of the most relevant candidate service content selected from the global service pool. For example, a target number of candidate service content can be extracted, and then the target number of candidate service content with the highest probability can be selected from the global service pool based on the probability distribution output by the model.

[0074] In this implementation, the network structure within the processing model performs nonlinear transformations and weighted calculations on these input features. It can automatically identify key feature combinations. For example, the processing model can discover that the combination of "Sichuan cuisine" preference and the "cooking instruction" scenario is strongly associated with services such as "recipe search" and "kitchenware purchase," rather than "restaurant reservation." The model calculates a relevance score for each possible service at the output layer, and then selects the top N (e.g., 5) services with the highest scores as candidate services for this recommendation.

[0075] In this embodiment, by employing a processing model to analyze and process service preference information and usage attribute information, intelligent and accurate identification of multiple candidate service contents can be achieved. This significantly improves the accuracy and diversity of candidate service contents, while also possessing strong generalization capabilities, enabling it to handle unseen new content and scenarios, and providing comprehensive and accurate data support for the final service recommendation.

[0076] S302: Obtain the target user's operational preference information for multiple candidate service contents.

[0077] In this embodiment, the step of obtaining the target user's operation preference information for multiple candidate service contents may specifically include: obtaining the service operation preference information of the target device, and extracting the operation preference information of multiple candidate service contents from the service operation preference information.

[0078] In this embodiment, service operation preference information reflects a user's preferences for various functional services of the target device. This service operation preference information can be obtained by statistical analysis of historical operation data for each functional service. For example, querying the user's operation preference log reveals the following service operation preference information: during the "end credits" period, the click-through rate for "trailer" type services is as high as 80%; for "shopping" type services, the historical click-through rate is 0%; and in the "suspense" scenario, the click-through rate for "original work reading" service is approximately 50%.

[0079] S303: Based on operation preference information and scenario information, determine multiple recommended service contents and their recommendation priorities from multiple candidate service contents.

[0080] In this embodiment, the terminal can filter out content that the user is clearly not interested in from multiple candidate service contents based on operation preference information. For example, given that the user never clicks on shopping services, "purchase of figurines of characters from this drama" can be removed from the candidate list, resulting in multiple recommended service contents for the user.

[0081] In this embodiment, the terminal can also calculate the score of each recommended service content based on operation preference information and scenario information, and sort them based on the scores to obtain the recommendation priority of multiple recommended service contents.

[0082] In specific implementation, the steps of determining the recommendation priority of multiple recommended service contents based on operation preference information and scenario information may include: determining a first score for the multiple recommended service contents based on operation preference information; determining a second score for the multiple recommended service contents based on scenario information; for any recommended service content, weighting the first and second scores to obtain a third score for that recommended service content; and determining the recommendation priority of the multiple recommended service contents based on the third scores. The second score can be calculated based on the matching degree between the scenario information and the multiple recommended service contents.

[0083] S304: Output multiple recommended service contents based on recommendation priority.

[0084] In this embodiment, the terminal can render and display multiple recommended service contents on the user interface based on a calculated priority order. These recommended service contents can be presented as service cards.

[0085] In this implementation, by introducing operational preference information and scenario information, the multiple candidate service contents obtained from the initial screening are further filtered and intelligently sorted. This proactively filters out content that users are not interested in and prioritizes high-value content, further improving the accuracy of service recommendations. This significantly improves user selection efficiency and click-through rate, thereby greatly enhancing user stickiness and satisfaction.

[0086] In this embodiment, after outputting multiple recommended service contents based on recommendation priority, the service recommendation method further includes: in response to an operation instruction for any of the multiple recommended service contents, executing the functional service corresponding to that recommended service content.

[0087] In this implementation, the operation command is feedback information triggered by the target user in response to recommended service content. For example, the target user can typically trigger the operation command via the confirmation button on a remote control or a touchscreen click. The smart TV can capture operation commands by monitoring focus changes and confirmation events on the user interface in real time.

[0088] In this implementation, each recommended service is bound to one or more executable functional services in the background. Upon receiving an operation instruction for any recommended service, the execution logic or service interface pre-associated with that recommended service will be immediately invoked to execute the functional service corresponding to that recommended service.

[0089] In this embodiment, by using the recommended service content as the entry point for the function service, a closed-loop interaction of intelligent recommendation can be achieved, thereby significantly improving interaction efficiency, reducing operating costs, and providing users with a seamless, smooth, and efficient user experience.

[0090] In one example, referring to Figure 4, a schematic diagram of the display interface for multiple recommended service contents is shown. A user is watching a movie starring actor A and actor B on a smart TV. During playback, the user triggers a key event by long-pressing the voice button on the remote control. The smart TV responds to this key event, recognizing the user's service preference information as a male fan of Hong Kong films, particularly enjoying art films starring actor A and crime films starring actor B. Simultaneously, through image recognition or metadata, the scene information is identified as the movie playback scene, and the main actors are actors A and B; the movie has been playing for approximately 2 minutes, and combined with the total content duration, the time period is determined to be the beginning of the movie. Then, the two sets of information, {service preference information: Hong Kong movies, actor A, actor B} and {usage attribute information: movie playback scene, opening sequence}, are input into a preset processing model. The processing model outputs a series of candidate service contents. Combining the user's operation preference information for the candidate service contents (for example, the user frequently uses the "skip the opening sequence" function at the beginning of movies and has a high click rate for "sound effects settings"), the multiple candidate service contents are filtered and sorted. Finally, based on the priority from high to low, four recommended service contents are obtained: "skip the opening sequence", "time reminder", "sound effects settings" and "graphic effects settings". Finally, the service cards corresponding to the four recommended service contents are rendered.

[0091] Subsequently, the user clicks the "Skip Intro" service card, and the smart TV responds to the click operation, identifies the intro position of the currently playing video, and performs the intro skip operation based on the intro position, positioning the video progress to the time point when the intro ends, thereby realizing the function service of skipping the intro with one click.

[0092] In this implementation, the user is ultimately presented with a collection of service cards that appear in batches, have clear priorities, and are highly related. This allows users to access the desired function with a single click without having to navigate through multiple menus. This significantly simplifies the operation path, enriches the interactive content, expands the functional boundaries of physical buttons, and effectively enhances the intelligence level of the service ecosystem.

[0093] In a feasible implementation, referring to FIG5, after outputting multiple recommended service contents based on recommendation priority, the service recommendation method may further include steps S401 to S403, as follows: S401: Record the operation result information of multiple recommended service contents.

[0094] In this embodiment, the operation result information of any recommended service content is used to characterize whether the target user has used the functional service corresponding to the recommended service content.

[0095] For example, a smart TV displays a set of recommended services to a target user, including: "Skip Intro," "Time Reminder," "Sound Effects Settings," and "Graphic Effects Settings." The user selects and executes "Skip Intro," "Sound Effects Settings," and "Graphic Effects Settings" in sequence using the remote control, and then ignores "Time Reminder." The smart TV records the operation result information of the above recommended services as: the user clicked and executed "Skip Intro," "Sound Effects Settings," and "Graphic Effects Settings," and ignored "Time Reminder."

[0096] S402: Update the operation preference information for multiple recommended service contents based on the operation result information.

[0097] In this embodiment, the operation result information of multiple recommended service contents can reflect the target user's operation preferences for multiple recommended service contents. Therefore, by updating the operation preference information of multiple recommended service contents through the operation result information, the accuracy of the operation preference information of recommended service contents can be continuously improved.

[0098] For example, if a user triggers the "Skip Intro" feature, its click-through rate is increased in the "Movie Intro" scenario. Conversely, the "Time Reminder" feature is ignored more frequently in the "Movie Intro" scenario, resulting in a decreased weight. The next time the smart TV generates recommendations for the same user in the same scenario, "Skip Intro" will have a higher priority, while "Time Reminder" may be ranked lower or even filtered out.

[0099] S403: Upload the operation result information to the cloud, and update the model parameters of the processed model based on the model update parameters returned by the cloud.

[0100] In this embodiment, the model update parameters are determined by training the cloud model based on the operation result information and by using the updated cloud model parameters.

[0101] In this embodiment, the cloud can receive encrypted operation result information from multiple terminals connected to it and aggregate it into a massive training dataset. The cloud can use this training dataset to train the cloud model (incremental training or full training) to obtain the updated model parameters of the cloud model, and calculate the model update parameters based on the updated model parameters.

[0102] In this embodiment, the model update parameter can represent the change in model parameters, which can be the difference between the model parameters of the cloud model and the model parameters of the processed model. That is, the cloud does not send the complete model parameters of the cloud model to the terminal, but only sends the change in model parameters to the terminal, so that the terminal updates the processed model according to the change in model parameters. In this way, the amount of data transmitted in each collaborative learning session can be greatly reduced.

[0103] In this embodiment, after receiving the model update parameters returned by the cloud, the terminal can update the parameters of the processing model deployed on the terminal, thereby achieving synchronization between the processing model and the cloud model.

[0104] In this embodiment, the model update parameter is a parameter used to update the model parameters of the processing model. For example, when the model update parameter uses the change in model parameters, the terminal can update the latest model parameters of the processing model by summing the change in model parameters and the model parameters of the processing model. Thus, by adopting an architecture combining edge computing and cloud-based federated learning, real-time inference and continuous optimization of the processing model can be achieved, improving the service recommendation accuracy of the processing model in various complex scenarios.

[0105] For example, during the training of a cloud model using a training dataset, the cloud model learns new global patterns, such as: "During the movie intro, the 'skip intro' service is widely needed, while the click-through rate of 'behind-the-scenes footage' is generally low." After updating the model parameters of the processing model based on the model update parameters returned by the cloud, since the cloud model has learned the above-mentioned correlation patterns from the global data, even when faced with a new user, the terminal will be more likely to prioritize recommending the "skip intro" service during the movie intro.

[0106] In this embodiment, by establishing the aforementioned update mechanism, closed-loop optimization of operation preference information and the processing model can be achieved. On the one hand, updating the operation preference information enables rapid adaptation to changes in user operating habits; on the other hand, by adopting an architecture combining edge computing and cloud-based federated learning, the accuracy and generalization ability of the processing model can be continuously optimized. This dual update mechanism effectively solves the model aging problem caused by user preference drift or content updates, while continuously improving the intelligence level of service recommendation.

[0107] To facilitate better implementation of the service recommendation method of this application, this application also provides a service recommendation apparatus based on the above-described service recommendation method. The meanings of the terms used are the same as in the service recommendation method described above, and specific implementation details can be found in the descriptions of the method embodiments.

[0108] Based on the same inventive concept, referring to FIG6, this application provides a service recommendation device 600, which includes: an information determination module 601, used to determine the service preference information of the target user and the usage attribute information of the target device in response to the current key event of the target device; wherein the usage attribute information includes scenario information and time period information; and a service recommendation module 602, used to output target service recommendation information based on the service preference information and the usage attribute information.

[0109] In one embodiment, the current key event includes key type information; the information determination module 601 includes: a user information acquisition submodule, used to acquire user feature information of the target user; and a service preference determination submodule, used to determine the service preference information of the target user based on the user feature information and key type information.

[0110] In one embodiment, the information determination module 601 includes: a content acquisition submodule for acquiring the currently playing content of the target device; and a content recognition submodule for recognizing the currently playing content to obtain scene information and time period information; wherein, the scene information represents the content scene information of the currently playing content, and the time period information represents the content time period information of the currently playing content.

[0111] In one embodiment, the target service recommendation information includes multiple recommended service contents; the service recommendation module 602 includes: a candidate content determination submodule, used to determine multiple candidate service contents based on service preference information and usage attribute information; a preference information acquisition submodule, used to acquire the target user's operation preference information for multiple candidate service contents; a recommendation content determination submodule, used to determine multiple recommended service contents and the recommendation priority of multiple recommended service contents among multiple candidate service contents based on operation preference information and scenario information; and a recommendation content output submodule, used to output multiple recommended service contents based on the recommendation priority.

[0112] In one embodiment, the service recommendation device 600 further includes a service execution module, configured to execute the functional service corresponding to any one of the multiple recommended service contents in response to an operation instruction for any one of the recommended service contents.

[0113] In one embodiment, the candidate content determination submodule is specifically used to input service preference information and usage attribute information into a preset processing model to obtain multiple candidate service contents output by the processing model.

[0114] In one embodiment, the service recommendation device 600 further includes: an operation recording module for recording operation result information of multiple recommended service contents; a preference update module for updating the operation preference information of multiple recommended service contents based on the operation result information; and a model update module for uploading the operation result information to the cloud and updating the model parameters of the processing model based on the model update parameters returned by the cloud; wherein the model update parameters are determined by the cloud based on the operation result information, training the cloud model, and the updated model parameters of the cloud model.

[0115] The technical solution adopted in this application determines the target user's service preference information and the target device's usage attribute information in response to the current key press event on the target device. The usage attribute information includes scenario information and time period information. Based on the service preference information and usage attribute information, target service recommendation information can be output. Thus, when a target user's key press operation is detected, user service preferences and device usage attributes can be automatically identified, achieving intelligent and accurate personalized service recommendations. This expands the functional boundaries of physical buttons while effectively improving the intelligence level of the service ecosystem, thereby significantly enhancing interaction efficiency and user experience.

[0116] Specific limitations regarding the service recommendation device 600 can be found in the limitations of the service recommendation method described above, and will not be repeated here. Each module in the aforementioned service recommendation device 600 can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0117] Furthermore, this application also provides an electronic device, as shown in FIG7, which illustrates the structural schematic diagram of the electronic device involved in this application. Specifically, the electronic device may include components such as a processor 701 with one or more processing cores and a memory 702 with one or more computer-readable storage media. Those skilled in the art will understand that the electronic device structure shown in FIG7 does not constitute a limitation on the electronic device, and may include more or fewer components than shown, or combine certain components, or have different component arrangements. The processor 701 is the control center of the electronic device, connecting various parts of the entire electronic device through various interfaces and lines. By running or executing software programs and / or modules stored in the memory 702, and calling data stored in the memory 702, it performs various functions of the electronic device and processes data, thereby performing overall monitoring of the electronic device. Optionally, the processor 701 may include one or more processing cores; preferably, the processor 701 may integrate an application processor and a modem processor, wherein the application processor mainly handles the operating system, user interface, and application programs, and the modem processor mainly handles wireless communication. It is understood that the aforementioned modem processor may also not be integrated into the processor 701.

[0118] The memory 702 can be used to store software programs and modules. The processor 701 executes various functional applications and data processing by running the software programs and modules stored in the memory 702. The memory 702 may mainly include a program storage area and a data storage area. The program storage area may store the operating system, application programs required for at least one function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created according to the use of the electronic device, etc. In addition, the memory 702 may include high-speed random access memory, and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other volatile solid-state storage device. Accordingly, the memory 702 may also include a memory controller to provide the processor 701 with access to the memory 702.

[0119] In one feasible implementation, the electronic device further includes a power supply 703 that supplies power to the various components. Preferably, the power supply 703 can be logically connected to the processor 701 through a power management system, thereby enabling functions such as charging, discharging, and power consumption management through the power management system. The power supply 703 may also include one or more DC or AC power supplies, recharging systems, power equipment debugging circuits, power converters or inverters, power status indicators, and other arbitrary components.

[0120] In one feasible implementation, the electronic device may further include an input unit 704, which can be used to receive input digital or character information and generate keyboard, mouse, joystick, optical or trackball signal inputs related to user settings and function control.

[0121] Although not shown, the electronic device may also include a display unit, etc., which will not be described in detail here. Specifically, in this embodiment, the processor 701 in the electronic device loads the executable files corresponding to the processes of one or more applications into the memory 702 according to the following instructions, and the processor 701 runs the applications stored in the memory 702, thereby implementing the steps in any of the service recommendation methods provided in the embodiments of this application.

[0122] Those skilled in the art will understand that the structure shown in Figure 7 is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the electronic device to which the present application is applied. The specific electronic device may include more or fewer components than shown in the figure, or combine certain components, or have different component arrangements.

[0123] In one feasible implementation, an electronic device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the methods described in any embodiment of this application.

[0124] In one feasible implementation, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the methods described in any embodiment of this application.

[0125] In one feasible implementation, a computer program product is also proposed, comprising a computer program or instructions that, when executed by a processor, implement the methods described in any embodiment of this application.

[0126] For details on the implementation of each of the above operations, please refer to the previous examples, which will not be repeated here.

[0127] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be performed by instructions, or by instructions controlling related hardware. These instructions can be stored in a computer-readable storage medium and loaded and executed by a processor.

[0128] Therefore, this application provides a computer-readable storage medium storing a computer program that can be loaded by a processor to perform the steps in any of the service recommendation methods provided in this application.

[0129] For details on the implementation of each of the above operations, please refer to the previous examples, which will not be repeated here.

[0130] The computer-readable storage medium may include: read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0131] Since the instructions stored in the computer-readable storage medium can execute the steps of any of the service recommendation methods provided in this application, the beneficial effects that any of the service recommendation methods provided in this application can achieve can be realized, as detailed in the preceding embodiments, and will not be repeated here.

[0132] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or terminal device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or terminal device. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or terminal device that includes said element.

[0133] The above provides a detailed description of a service recommendation method, apparatus, electronic device, and computer-readable storage medium provided in this application. Specific examples have been used to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only for the purpose of helping to understand the method and core ideas of the present invention. At the same time, those skilled in the art will recognize that there will be changes in the specific implementation methods and application scope based on the ideas of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A service recommendation method, characterized in that, The method includes: in response to a current key event on a target device, determining service preference information of a target user and determining usage attribute information of the target device; wherein the usage attribute information includes scenario information and time period information; and outputting target service recommendation information based on the service preference information and the usage attribute information.

2. The service recommendation method according to claim 1, characterized in that, The current key event includes key type information; determining the target user's service preference information includes: obtaining the target user's user characteristic information; and determining the target user's service preference information based on the user characteristic information and the key type information.

3. The service recommendation method according to claim 1, characterized in that, The step of determining the usage attribute information of the target device includes: obtaining the currently playing content of the target device; identifying the currently playing content to obtain the scene information and the time period information; wherein, the scene information represents the content scene information of the currently playing content, and the time period information represents the content time period information of the currently playing content.

4. The service recommendation method according to claim 1, characterized in that, The target service recommendation information includes multiple recommended service contents; The step of outputting target service recommendation information based on the service preference information and the usage attribute information includes: determining multiple candidate service contents based on the service preference information and the usage attribute information; obtaining the target user's operation preference information for the multiple candidate service contents; determining multiple recommended service contents and their recommendation priorities among the multiple candidate service contents based on the operation preference information and the scenario information; and outputting the multiple recommended service contents based on the recommendation priorities.

5. The service recommendation method according to claim 4, characterized in that, After outputting multiple recommended service contents based on the recommendation priority, the method further includes: in response to an operation instruction for any of the multiple recommended service contents, executing the functional service corresponding to that recommended service content.

6. The service recommendation method according to claim 4, characterized in that, The step of determining multiple candidate service contents based on the service preference information and the usage attribute information includes: inputting the service preference information and the usage attribute information into a preset processing model to obtain multiple candidate service contents output by the processing model.

7. The service recommendation method according to claim 6, characterized in that, After outputting multiple recommended service contents based on the recommendation priority, the method further includes: recording operation result information of the multiple recommended service contents; updating operation preference information of the multiple recommended service contents based on the operation result information; and / or uploading the operation result information to the cloud and updating the model parameters of the processing model based on the model update parameters returned by the cloud; wherein the model update parameters are determined by the cloud to train the cloud model based on the operation result information and based on the updated model parameters of the cloud model.

8. A service recommendation device, characterized in that, The device includes: an information determination module, configured to determine the service preference information of the target user and the usage attribute information of the target device in response to a current key event on the target device; wherein the usage attribute information includes scenario information and time period information; and a service recommendation module, configured to output target service recommendation information based on the service preference information and the usage attribute information.

9. An electronic device, characterized in that, It includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the service 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 a computer program that, when executed by a processor, implements the steps of the service recommendation method as described in any one of claims 1 to 7.

Images