Method and device for scene recommendation in smart home system

Abstract

The embodiment of the invention provides a scene recommendation method and device in a smart home system, and is applied to the technical field of smart home, and the method comprises the steps: an edge device in the smart home system receives a first scene recommendation list sent by a cloud server; the first scene recommendation list comprises information of a plurality of intelligent scenes, and the information of the intelligent scenes is used for controlling intelligent home equipment in the intelligent home system to execute corresponding operation; according to local data of the smart home system, performing personalized adjustment on the first scene recommendation list to obtain a second scene recommendation list; and performing intelligent scene recommendation according to the second scene recommendation list. According to the embodiment of the invention, the edge device in the smart home system performs personalized adjustment on the basis of the smart scene recommended by the cloud server, and the effect of smart scene recommendation is improved.

Description

Technical Field

[0001] This invention relates to the field of smart home technology, and in particular to a method and apparatus for scene recommendation in a smart home system. Background Technology

[0002] With the development of IoT technology, smart home devices are becoming increasingly complex and widely used. In daily use, it's often necessary to recommend suitable smart scenarios to users, who then select and apply them. However, current smart scenario recommendations are delivered directly to users by cloud servers, resulting in ineffective recommendations. Summary of the Invention

[0003] In view of the above problems, a method and apparatus for scene recommendation in a smart home system are proposed to overcome or at least partially solve the above problems, including:

[0004] A method for scene recommendation in a smart home system, the method comprising:

[0005] The system receives a first scene recommendation list sent by a cloud server; the first scene recommendation list includes information on multiple smart scenes, and the information on the smart scenes is used to control the smart home devices in the smart home system to perform corresponding operations.

[0006] Based on the local data of the smart home system, the first scene recommendation list is personalized to obtain the second scene recommendation list;

[0007] Based on the second scenario recommendation list, intelligent scenario recommendations are made.

[0008] Optionally, the first scene recommendation list can be personalized and adjusted according to the local data of the smart home system to obtain a second scene recommendation list, including: reordering multiple smart scenes included in the first scene recommendation list according to the local data of the smart home system to obtain the second scene recommendation list.

[0009] Optionally, based on local data from the smart home system, the multiple smart scenes included in the first scene recommendation list are reordered to obtain a second scene recommendation list, including:

[0010] When a change in local data of the smart home system is detected, information about changes in user intent or environment in the smart home system is determined.

[0011] Based on the user intent change information or environment change information, the multiple intelligent scenarios included in the first scenario recommendation list are reordered to obtain the second scenario recommendation list.

[0012] Optionally, based on local data from the smart home system, the multiple smart scenes included in the first scene recommendation list are reordered to obtain a second scene recommendation list, including:

[0013] Determine user negative feedback data from the local data of the smart home system;

[0014] Based on the user negative feedback data, the multiple intelligent scenes included in the first scene recommendation list are reordered to obtain the second scene recommendation list.

[0015] Optionally, the user negative feedback data includes: relevant data that was recommended to the user but was not selected by the user.

[0016] Optionally, before reordering the multiple smart scenes included in the first scene recommendation list based on the local data of the smart home system to obtain the second scene recommendation list, the method further includes:

[0017] Based on the local data and the cloud data from the cloud server, determine the differences;

[0018] From the multiple intelligent scenarios included in the first scenario recommendation list, a target intelligent scenario corresponding to the difference item is determined, and the information of the target intelligent scenario is adjusted.

[0019] Optionally, the method, when applied to an edge device, further includes:

[0020] When a user logs into the smart home system for the first time, the user's relevant data is copied from the edge device of the smart home system in which the user previously logged in.

[0021] A scene recommendation device in a smart home system, the device comprising:

[0022] The first scene recommendation list receiving module is used to receive a first scene recommendation list sent by the cloud server; the first scene recommendation list includes information on multiple smart scenes, and the information on the smart scenes is used to control the smart home devices in the smart home system to perform corresponding operations.

[0023] The second scenario recommendation list acquisition module is used to personalize the first scenario recommendation list based on the local data of the smart home system to obtain the second scenario recommendation list;

[0024] The intelligent scene recommendation module is used to make intelligent scene recommendations based on the second scene recommendation list.

[0025] An electronic device includes a processor, a memory, and a computer program stored in the memory and capable of running on the processor, wherein the computer program, when executed by the processor, implements the method described above.

[0026] A computer-readable storage medium on which a computer program is stored, which, when executed by a processor, implements the method described above.

[0027] A computer program product includes a computer program that, when executed by a processor, implements the method described above.

[0028] The embodiments of the present invention have the following advantages:

[0029] In this embodiment of the invention, edge devices in a smart home system receive a first scene recommendation list sent by a cloud server. The first scene recommendation list includes information on multiple smart scenes, and the information on the smart scenes is used to control the smart home devices in the smart home system to perform corresponding operations. Then, based on the local data of the smart home system, the first scene recommendation list is personalized to obtain a second scene recommendation list. Based on the second scene recommendation list, smart scene recommendations are made. This realizes that the edge devices in the smart home system can make personalized adjustments based on the smart scenes recommended by the cloud server, thereby improving the effect of smart scene recommendations. Attached Figure Description

[0030] To more clearly illustrate the technical solution of the present invention, the accompanying drawings used in the description of the present invention will be briefly introduced below. Obviously, the accompanying 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.

[0031] Figure 1 This is a flowchart illustrating the steps of a scene recommendation method in a smart home system according to some embodiments of the present invention;

[0032] Figure 2 This is a flowchart of the steps of another method for scene recommendation in a smart home system provided by some embodiments of the present invention;

[0033] Figure 3 This is a flowchart of the steps of another method for scene recommendation in a smart home system provided by some embodiments of the present invention;

[0034] Figure 4 This is a flowchart of the steps of another method for scene recommendation in a smart home system provided by some embodiments of the present invention;

[0035] Figure 5This is a flowchart of the steps of another method for scene recommendation in a smart home system provided by some embodiments of the present invention;

[0036] Figure 6 This is a structural block diagram of another scene recommendation device in a smart home system provided by some embodiments of the present invention. Detailed Implementation

[0037] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0038] In the process of intelligent scene recommendation, cloud servers can combine various information such as long-term received device status information, user information, and environmental information, and use big data analysis to build large-scale model algorithms to select and rank intelligent scenes. Through long-term, large-scale data iteration, updates, and adjustments, cloud servers have become quite mature in terms of the breadth of their models.

[0039] However, running models on cloud servers incurs huge cloud resource overhead, and since cloud servers output results after analyzing large amounts of data, data generated by users in real time will only be reflected in the push notifications after a considerable delay, making it impossible to make refined recommendations based on users' current behavior.

[0040] To improve the understanding of user intent in different scenarios and environments, and reduce the update latency of smart scene recommendation list information and the perception latency of changes in real-time device and environmental information, the recommendation information can be processed again on edge devices in the smart home system to achieve cloud-edge collaborative optimization.

[0041] To achieve cloud-edge collaborative optimization, a cloud coarse screening layer and an edge fine sorting layer can be set.

[0042] The cloud-based coarse screening layer aggregates global user behavior data (such as APP click data, smart home operation data, and wearable device data) through a multi-task network to construct a heterogeneous interest representation tower (Home Tower) to better understand users' interests in different areas. Furthermore, by introducing a scene selection network, a cross-domain mapping matrix is ​​automatically generated from home environment (such as temperature, humidity, and lighting) to recommended scenes (such as comfort and energy saving). Subsequently, high-frequency, high-value scene data, such as commonly used scenes, are manually weighted.

[0043] The edge fine-tuning layer selects a lightweight version of the network, fine-tuning it based on device status data streams, user control, and voice commands. It employs a negative feedback mechanism, introducing multiple forgetting mechanisms, such as allowing users to skip certain commands, removing repetitive scenarios, and removing data from areas that have not been used for a long time. Furthermore, if a user logs in from a different location, features can be quickly copied to another edge node by directly transmitting model features.

[0044] In some examples, M / M / s (Poisson Arrival, Exponential Service, s Servers Queueing Model) queuing theory can be applied between cloud servers and edge devices to optimize resource allocation and task scheduling, ensuring efficient operation.

[0045] In some examples, the number of service nodes and resource allocation strategies can be dynamically adjusted based on user behavior and changes in the scenario to improve system response speed and user experience. For instance, if a particular recommendation feature is frequently recommended and used, the resource allocation to the servers used for that feature can be increased.

[0046] The present invention will be further described below with reference to the accompanying drawings:

[0047] Reference Figure 1 This diagram illustrates a flowchart of the steps of a scene recommendation method in a smart home system according to some embodiments of the present invention. This method can be applied to edge devices in a smart home system. Edge devices are devices located in the local environment of the smart home system, such as central control devices, smart speakers, and smart routers. In some examples, edge devices may also include home host systems, home servers, and other similar devices.

[0048] Specifically, it may include the following steps:

[0049] Step 101: Receive a first scene recommendation list sent by the cloud server; the first scene recommendation list includes information on multiple smart scenes, and the information on the smart scenes is used to control the smart home devices in the smart home system to perform corresponding operations.

[0050] As an example, smart scenes can be movie viewing mode, away from home mode, home mode, sleep mode, etc. The information in these smart scenes includes instructions or parameters to control smart home devices to perform corresponding operations. For example, in movie viewing mode, the lights can be dimmed automatically, the curtains can be closed, and the air conditioner temperature can be adjusted.

[0051] In practical applications, such as Figure 2The cloud server can combine various information such as long-term received device status information, user information, environmental information, positive feedback information, and negative feedback information, and use big data analysis to build a large-scale model algorithm to rank intelligent scenarios and periodically generate a first scenario recommendation list. Edge devices receive the first scenario recommendation list from the cloud server.

[0052] In some examples, cloud servers sort smart scenes based on big data analysis results, according to factors such as popularity, user preferences, and frequency of scene use, and generate a first scene recommendation list containing information on multiple smart scenes. This list not only includes the names of the smart scenes, but also the specific instructions or parameters required to execute these scenes, ensuring that smart home devices can accurately perform the corresponding operations.

[0053] In some examples, to achieve cloud-edge collaborative optimization, a cloud-based coarse screening layer and an edge-based fine screening layer can be set up. The cloud-based coarse screening layer aggregates global user behavior data (such as APP click data, smart home operation data, and wearable device data) through a multi-task network to construct a heterogeneous interest representation tower (Home Tower) to better understand users' interests in different domains. Moreover, by introducing a scene selection network, a cross-domain mapping matrix from home environment (such as temperature, humidity, and lighting) to recommended scenes (such as comfort and energy saving) is automatically generated. Subsequently, high-frequency and high-value scene data, such as some commonly used scenes, are manually weighted.

[0054] Step 102: Based on the local data of the smart home system, the first scene recommendation list is personalized to obtain the second scene recommendation list.

[0055] After obtaining the initial scenario recommendation list from the cloud server, edge devices can further filter and optimize the smart scenarios in the recommendation list based on local data from the smart home system. In some examples, local data may include, but is not limited to, the real-time status of smart home devices, the user's historical operation records, and current environmental conditions. By analyzing this data, edge devices can more accurately understand the user's current needs and preferences, thereby personalizing the initial scenario recommendation list.

[0056] In some examples, the personalization process can employ various algorithms and strategies. For instance, edge devices can statistically analyze the frequency of smart scene usage based on the user's historical operation records, prioritizing frequently used scenes to improve recommendation accuracy. Simultaneously, edge devices can intelligently match smart scenes based on current environmental conditions, such as time, temperature, and humidity, recommending the most suitable scenes for the current environment. After personalization, the edge device obtains a second scene recommendation list. Compared to the first scene recommendation list, the second list better matches the user's current needs and preferences, thus providing a higher-quality smart scene recommendation service.

[0057] In some examples, to achieve cloud-edge collaborative optimization, a cloud-based coarse screening layer and an edge-based fine ranking layer can be configured. The edge fine ranking layer selects a lightweight version of the network, fine-tunes it based on device status data streams, user controls, and voice commands, and employs a negative feedback mechanism, introducing various forgetting mechanisms, such as allowing users to skip certain commands, removing repetitive scenarios, and removing data from areas that have not been used for a long time. Furthermore, if a user logs in from a different location, features can be quickly copied to another edge node by directly transmitting model features.

[0058] In this embodiment of the invention, the local model of the smart home system can perform calculations based on only a small amount of local data (such as local real-time status or operating parameters) to personalize the first scene recommendation list, thereby significantly optimizing and reducing the computational load and size of the model.

[0059] In some embodiments of the present invention, the first scene recommendation list is personalized based on local data of the smart home system to obtain a second scene recommendation list, including:

[0060] Based on the local data of the smart home system, the multiple smart scenes included in the first scene recommendation list are reordered to obtain the second scene recommendation list.

[0061] In practical applications, edge devices can reorder the order of multiple smart scenes in the initial recommendation list based on local data from the smart home system, prioritizing those scenes that better match the user's current needs and preferences. For example, if a user frequently watches movies at night, the edge device can prioritize the movie-watching mode for easier selection and use. In some examples, the reordering process can employ various algorithms and strategies, such as frequency statistics based on the user's historical operation records or intelligent matching based on current environmental conditions, to ensure the accuracy and intelligence of the recommendations.

[0062] In some embodiments of the present invention, the multiple smart scenes included in the first scene recommendation list are reordered according to the local data of the smart home system to obtain the second scene recommendation list, including: when a change in the local data of the smart home system is detected, determining user intent change information or environment change information in the smart home system; and reordering the multiple smart scenes included in the first scene recommendation list according to the user intent change information or environment change information to obtain the second scene recommendation list.

[0063] When changes are detected in local data such as the current environment, operation, and status of the smart home system, such as Figure 2 If the user's current environment or intent has changed, then the user intent change information or environment change information can be determined. Then, based on the user intent change information or environment change information, the corresponding local end arrangement scene can be updated in real time, and the multiple intelligent scenes contained in the first scene recommendation list can be reordered to realize the recommendation list change under real-time conditions.

[0064] In some examples, when local data changes, it can be uploaded to a cloud server, which then optimizes the model based on the new data.

[0065] In some embodiments of the present invention, the process of reordering multiple smart scenes included in the first scene recommendation list to obtain a second scene recommendation list based on local data of the smart home system includes: determining user negative feedback data from the local data of the smart home system; and reordering multiple smart scenes included in the first scene recommendation list based on the user negative feedback data to obtain the second scene recommendation list.

[0066] In some examples, negative user feedback data includes data that was recommended to the user but was not selected by the user, as well as data that the user skipped command execution, data from repetitive scenarios, and data that has not been used for a long time.

[0067] In practical applications, cloud server algorithms primarily use positive feedback mechanisms such as clicks or traffic to calculate and rank content; more selections result in higher exposure and a higher ranking. On the edge device side, however, negative feedback can be used to acquire user feedback data, which is then used to reorder the multiple intelligent scenarios within the initial recommendation list.

[0068] For example, the concept of a parameter that is displayed but not selected is introduced (i.e., relevant data recommended to users but not selected by them). When a user receives a corresponding push notification but does not select an option, the corresponding priority is reduced when subsequent push notifications are rearranged at the edge.

[0069] In some embodiments of the present invention, it further includes:

[0070] When a user logs into the smart home system for the first time, the user's relevant data is copied from the edge device of the smart home system in which the user previously logged in.

[0071] In scenarios such as abnormal logins or when a user logs into a smart home system for the first time, the user's relevant data can be copied from the edge device of the smart home system where the user first logged in and applied to the local edge device. This allows for the rapid replication of features at another edge node by directly transmitting model features.

[0072] In some embodiments of the present invention, before reordering the multiple smart scenes included in the first scene recommendation list based on the local data of the smart home system to obtain the second scene recommendation list, the method further includes: determining the difference items based on the local data and the cloud data of the cloud server; determining the target smart scene corresponding to the difference items from the multiple smart scenes included in the first scene recommendation list, and adjusting the information of the target smart scene.

[0073] In practical applications, there may be some discrepancies between the data from edge devices and cloud servers. For example, local data may be more real-time and granular, while cloud data is more comprehensive and macroscopic. To ensure that the recommended smart scenarios not only meet the user's current needs but also make full use of cloud server resources, the differences between local and cloud data can be identified before reordering. These differences may include changes in user preferences, updates to the status of smart home devices, and differences in environmental conditions.

[0074] After identifying the discrepancies, target intelligent scenarios corresponding to these discrepancies can be determined from the multiple intelligent scenarios included in the first scenario recommendation list, and the information of these target intelligent scenarios can be adjusted. For example, if local data shows that a user's preference for a certain intelligent scenario has changed, the position of that scenario in the recommendation list can be adjusted, or the specific instructions or parameters executed can be modified. Through such adjustments, it can be ensured that the recommended intelligent scenarios better match the user's current needs and preferences, improving the accuracy and intelligence level of the recommendations.

[0075] For example, the difference could be temperature. When there's a discrepancy between the temperature value in local data and the temperature value in cloud data, if the local data shows a higher temperature, the user might be more inclined to use cooling devices like air conditioners or fans. Therefore, cooling-related smart scenarios (such as "Cooling Mode") could be prioritized in the recommendation list. Conversely, if the local data shows a lower temperature, the user might be more inclined to use heating devices like heaters. Thus, heating-related smart scenarios (such as "Warm Mode") could be prioritized in the recommendation list. This adjustment ensures that the recommended smart scenarios are more closely matched to the user's current environmental conditions and needs.

[0076] Step 103: Perform intelligent scene recommendation based on the second scene recommendation list.

[0077] After obtaining the second scenario recommendation list, edge devices can recommend smart scenarios to users based on this list. For example, the recommendation list can be displayed through the central control device of the smart home system, and users can select the desired scenario by touching the screen; alternatively, smart speakers can read the recommendation list aloud, and users can select scenarios by voice commands. In practical applications, edge devices can also combine the real-time status of smart home devices and the user's historical operation records to intelligently predict the user's next needs and proactively recommend corresponding smart scenarios, further enhancing the user's smart home experience.

[0078] In some examples, during the intelligent scene recommendation process, edge devices can collect user feedback data in real time, such as the user's acceptance of the recommended scenes and the frequency of use, and upload this data to the cloud server. The cloud server can then continuously optimize the model algorithm based on this feedback data to improve the accuracy and intelligence of the intelligent scene recommendations. Through this cloud-edge collaborative optimization mechanism, it can be ensured that the smart home system can consistently provide intelligent scene recommendation services that meet user needs and preferences.

[0079] In this embodiment of the invention, edge devices in a smart home system receive a first scene recommendation list sent by a cloud server. The first scene recommendation list includes information on multiple smart scenes, and the information on the smart scenes is used to control the smart home devices in the smart home system to perform corresponding operations. Then, based on the local data of the smart home system, the first scene recommendation list is personalized to obtain a second scene recommendation list. Based on the second scene recommendation list, smart scene recommendations are made. This realizes that the edge devices in the smart home system can make personalized adjustments based on the smart scenes recommended by the cloud server, thereby improving the effect of smart scene recommendations.

[0080] Reference Figure 3The diagram illustrates a flowchart of another method for scene recommendation in a smart home system provided by some embodiments of the present invention, which may specifically include the following steps:

[0081] Step 301: Receive a first scene recommendation list sent by the cloud server; the first scene recommendation list includes information on multiple smart scenes, and the information on the smart scenes is used to control the smart home devices in the smart home system to perform corresponding operations.

[0082] As an example, smart scenes can be movie viewing mode, away from home mode, home mode, sleep mode, etc. The information in these smart scenes includes instructions or parameters to control smart home devices to perform corresponding operations. For example, in movie viewing mode, the lights can be dimmed automatically, the curtains can be closed, and the air conditioner temperature can be adjusted.

[0083] In practical applications, such as Figure 2 The cloud server can combine various information such as long-term received device status information, user information, environmental information, positive feedback information, and negative feedback information, and use big data analysis to build a large-scale model algorithm to rank intelligent scenarios and periodically generate a first scenario recommendation list. Edge devices receive the first scenario recommendation list from the cloud server.

[0084] In some examples, cloud servers sort smart scenes based on big data analysis results, according to factors such as popularity, user preferences, and frequency of scene use, and generate a first scene recommendation list containing information on multiple smart scenes. This list not only includes the names of the smart scenes, but also the specific instructions or parameters required to execute these scenes, ensuring that smart home devices can accurately perform the corresponding operations.

[0085] In some examples, to achieve cloud-edge collaborative optimization, a cloud-based coarse screening layer and an edge-based fine screening layer can be set up. The cloud-based coarse screening layer aggregates global user behavior data (such as APP click data, smart home operation data, and wearable device data) through a multi-task network to construct a heterogeneous interest representation tower (Home Tower) to better understand users' interests in different domains. Moreover, by introducing a scene selection network, a cross-domain mapping matrix from home environment (such as temperature, humidity, and lighting) to recommended scenes (such as comfort and energy saving) is automatically generated. Subsequently, high-frequency and high-value scene data, such as some commonly used scenes, are manually weighted.

[0086] Step 302: When a change in the local data of the smart home system is detected, determine the user intent change information or environment change information in the smart home system.

[0087] Step 303: Based on the user intent change information or environment change information, reorder the multiple smart scenes included in the first scene recommendation list to obtain the second scene recommendation list.

[0088] When changes are detected in local data such as the current environment, operation, and status of the smart home system, such as Figure 2 If the user's current environment or intent has changed, then the user intent change information or environment change information can be determined. Then, based on the user intent change information or environment change information, the corresponding local end arrangement scene can be updated in real time, and the multiple intelligent scenes contained in the first scene recommendation list can be reordered to realize the recommendation list change under real-time conditions.

[0089] In some examples, when local data changes, it can be uploaded to a cloud server, which then optimizes the model based on the new data.

[0090] Step 304: Perform intelligent scene recommendation based on the second scene recommendation list.

[0091] After obtaining the second scenario recommendation list, edge devices can recommend smart scenarios to users based on this list. For example, the recommendation list can be displayed through the central control device of the smart home system, and users can select the desired scenario by touching the screen; alternatively, smart speakers can read the recommendation list aloud, and users can select scenarios by voice commands. In practical applications, edge devices can also combine the real-time status of smart home devices and the user's historical operation records to intelligently predict the user's next needs and proactively recommend corresponding smart scenarios, further enhancing the user's smart home experience.

[0092] In some examples, during the intelligent scene recommendation process, edge devices can collect user feedback data in real time, such as the user's acceptance of the recommended scenes and the frequency of use, and upload this data to the cloud server. The cloud server can then continuously optimize the model algorithm based on this feedback data to improve the accuracy and intelligence of the intelligent scene recommendations. Through this cloud-edge collaborative optimization mechanism, it can be ensured that the smart home system can consistently provide intelligent scene recommendation services that meet user needs and preferences.

[0093] Reference Figure 4 The diagram illustrates a flowchart of another method for scene recommendation in a smart home system provided by some embodiments of the present invention, which may specifically include the following steps:

[0094] Step 401: Receive a first scene recommendation list sent by the cloud server; the first scene recommendation list includes information on multiple smart scenes, and the information on the smart scenes is used to control the smart home devices in the smart home system to perform corresponding operations.

[0095] As an example, smart scenes can be movie viewing mode, away from home mode, home mode, sleep mode, etc. The information in these smart scenes includes instructions or parameters to control smart home devices to perform corresponding operations. For example, in movie viewing mode, the lights can be dimmed automatically, the curtains can be closed, and the air conditioner temperature can be adjusted.

[0096] In practical applications, such as Figure 2 The cloud server can combine various information such as long-term received device status information, user information, environmental information, positive feedback information, and negative feedback information, and use big data analysis to build a large-scale model algorithm to rank intelligent scenarios and periodically generate a first scenario recommendation list. Edge devices receive the first scenario recommendation list from the cloud server.

[0097] In some examples, cloud servers sort smart scenes based on big data analysis results, according to factors such as popularity, user preferences, and frequency of scene use, and generate a first scene recommendation list containing information on multiple smart scenes. This list not only includes the names of the smart scenes, but also the specific instructions or parameters required to execute these scenes, ensuring that smart home devices can accurately perform the corresponding operations.

[0098] In some examples, to achieve cloud-edge collaborative optimization, a cloud-based coarse screening layer and an edge-based fine screening layer can be set up. The cloud-based coarse screening layer aggregates global user behavior data (such as APP click data, smart home operation data, and wearable device data) through a multi-task network to construct a heterogeneous interest representation tower (Home Tower) to better understand users' interests in different domains. Moreover, by introducing a scene selection network, a cross-domain mapping matrix from home environment (such as temperature, humidity, and lighting) to recommended scenes (such as comfort and energy saving) is automatically generated. Subsequently, high-frequency and high-value scene data, such as some commonly used scenes, are manually weighted.

[0099] Step 402: Determine user negative feedback data from the local data of the smart home system.

[0100] Step 403: Based on the user negative feedback data, reorder the multiple intelligent scenes included in the first scene recommendation list to obtain the second scene recommendation list.

[0101] In some examples, negative user feedback data includes data that was recommended to the user but was not selected by the user, as well as data that the user skipped command execution, data from repetitive scenarios, and data that has not been used for a long time.

[0102] In practical applications, cloud server algorithms primarily use positive feedback mechanisms such as clicks or traffic to calculate and rank content; more selections result in higher exposure and a higher ranking. On the edge device side, however, negative feedback can be used to acquire user feedback data, which is then used to reorder the multiple intelligent scenarios within the initial recommendation list.

[0103] For example, the concept of a parameter that is displayed but not selected is introduced (i.e., relevant data recommended to users but not selected by them). When a user receives a corresponding push notification but does not select an option, the corresponding priority is reduced when subsequent push notifications are rearranged at the edge.

[0104] Step 404: Perform intelligent scene recommendation based on the second scene recommendation list.

[0105] After obtaining the second scenario recommendation list, edge devices can recommend smart scenarios to users based on this list. For example, the recommendation list can be displayed through the central control device of the smart home system, and users can select the desired scenario by touching the screen; alternatively, smart speakers can read the recommendation list aloud, and users can select scenarios by voice commands. In practical applications, edge devices can also combine the real-time status of smart home devices and the user's historical operation records to intelligently predict the user's next needs and proactively recommend corresponding smart scenarios, further enhancing the user's smart home experience.

[0106] In some examples, during the intelligent scene recommendation process, edge devices can collect user feedback data in real time, such as the user's acceptance of the recommended scenes and the frequency of use, and upload this data to the cloud server. The cloud server can then continuously optimize the model algorithm based on this feedback data to improve the accuracy and intelligence of the intelligent scene recommendations. Through this cloud-edge collaborative optimization mechanism, it can be ensured that the smart home system can consistently provide intelligent scene recommendation services that meet user needs and preferences.

[0107] Reference Figure 5 The diagram illustrates a flowchart of another method for scene recommendation in a smart home system provided by some embodiments of the present invention, which may specifically include the following steps:

[0108] Step 501: Receive a first scene recommendation list sent by the cloud server; the first scene recommendation list includes information on multiple smart scenes, and the information on the smart scenes is used to control the smart home devices in the smart home system to perform corresponding operations.

[0109] As an example, smart scenes can be movie viewing mode, away from home mode, home mode, sleep mode, etc. The information in these smart scenes includes instructions or parameters to control smart home devices to perform corresponding operations. For example, in movie viewing mode, the lights can be dimmed automatically, the curtains can be closed, and the air conditioner temperature can be adjusted.

[0110] In practical applications, such as Figure 2 The cloud server can combine various information such as long-term received device status information, user information, environmental information, positive feedback information, and negative feedback information, and use big data analysis to build a large-scale model algorithm to rank intelligent scenarios and periodically generate a first scenario recommendation list. Edge devices receive the first scenario recommendation list from the cloud server.

[0111] In some examples, cloud servers sort smart scenes based on big data analysis results, according to factors such as popularity, user preferences, and frequency of scene use, and generate a first scene recommendation list containing information on multiple smart scenes. This list not only includes the names of the smart scenes, but also the specific instructions or parameters required to execute these scenes, ensuring that smart home devices can accurately perform the corresponding operations.

[0112] In some examples, to achieve cloud-edge collaborative optimization, a cloud-based coarse screening layer and an edge-based fine screening layer can be set up. The cloud-based coarse screening layer aggregates global user behavior data (such as APP click data, smart home operation data, and wearable device data) through a multi-task network to construct a heterogeneous interest representation tower (Home Tower) to better understand users' interests in different domains. Moreover, by introducing a scene selection network, a cross-domain mapping matrix from home environment (such as temperature, humidity, and lighting) to recommended scenes (such as comfort and energy saving) is automatically generated. Subsequently, high-frequency and high-value scene data, such as some commonly used scenes, are manually weighted.

[0113] Step 502: Determine the discrepancies based on the local data of the smart home system and the cloud data of the cloud server.

[0114] In practical applications, there may be some discrepancies between the data from edge devices and cloud servers. For example, local data may be more real-time and granular, while cloud data is more comprehensive and macroscopic. To ensure that the recommended smart scenarios not only meet the user's current needs but also make full use of cloud server resources, the differences between local and cloud data can be identified before reordering. These differences may include changes in user preferences, updates to the status of smart home devices, and differences in environmental conditions.

[0115] Step 503: From the multiple intelligent scenes included in the first scene recommendation list, determine the target intelligent scene corresponding to the difference item, and adjust the information of the target intelligent scene.

[0116] After identifying the discrepancies, target intelligent scenarios corresponding to these discrepancies can be determined from the multiple intelligent scenarios included in the first scenario recommendation list, and the information of these target intelligent scenarios can be adjusted. For example, if local data shows that a user's preference for a certain intelligent scenario has changed, the position of that scenario in the recommendation list can be adjusted, or the specific instructions or parameters executed can be modified. Through such adjustments, it can be ensured that the recommended intelligent scenarios better match the user's current needs and preferences, improving the accuracy and intelligence level of the recommendations.

[0117] For example, the difference could be temperature. When there's a discrepancy between the temperature value in local data and the temperature value in cloud data, if the local data shows a higher temperature, the user might be more inclined to use cooling devices like air conditioners or fans. Therefore, cooling-related smart scenarios (such as "Cooling Mode") could be prioritized in the recommendation list. Conversely, if the local data shows a lower temperature, the user might be more inclined to use heating devices like heaters. Thus, heating-related smart scenarios (such as "Warm Mode") could be prioritized in the recommendation list. This adjustment ensures that the recommended smart scenarios are more closely matched to the user's current environmental conditions and needs.

[0118] Step 504: Based on the local data of the smart home system, reorder the multiple smart scenes included in the first scene recommendation list to obtain the second scene recommendation list.

[0119] In practical applications, edge devices can reorder the order of multiple smart scenes in the initial recommendation list based on local data from the smart home system, prioritizing those scenes that better match the user's current needs and preferences. For example, if a user frequently watches movies at night, the edge device can prioritize the movie-watching mode for easier selection and use. In some examples, the reordering process can employ various algorithms and strategies, such as frequency statistics based on the user's historical operation records or intelligent matching based on current environmental conditions, to ensure the accuracy and intelligence of the recommendations.

[0120] Step 505: Perform intelligent scene recommendation based on the second scene recommendation list.

[0121] After obtaining the second scenario recommendation list, edge devices can recommend smart scenarios to users based on this list. For example, the recommendation list can be displayed through the central control device of the smart home system, and users can select the desired scenario by touching the screen; alternatively, smart speakers can read the recommendation list aloud, and users can select scenarios by voice commands. In practical applications, edge devices can also combine the real-time status of smart home devices and the user's historical operation records to intelligently predict the user's next needs and proactively recommend corresponding smart scenarios, further enhancing the user's smart home experience.

[0122] In some examples, during the intelligent scene recommendation process, edge devices can collect user feedback data in real time, such as the user's acceptance of the recommended scenes and the frequency of use, and upload this data to the cloud server. The cloud server can then continuously optimize the model algorithm based on this feedback data to improve the accuracy and intelligence of the intelligent scene recommendations. Through this cloud-edge collaborative optimization mechanism, it can be ensured that the smart home system can consistently provide intelligent scene recommendation services that meet user needs and preferences.

[0123] It should be noted that, for the sake of simplicity, the method embodiments are all described as a series of actions. However, those skilled in the art should understand that the embodiments of the present invention are not limited to the described order of actions, because according to the embodiments of the present invention, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions involved are not necessarily essential to the embodiments of the present invention.

[0124] Reference Figure 6 The diagram illustrates a structural schematic of a scene recommendation device in a smart home system provided by some embodiments of the present invention, which is applied to edge devices in a smart home system.

[0125] Specifically, it can include the following modules:

[0126] The first scene recommendation list receiving module 601 is used to receive a first scene recommendation list sent by a cloud server; the first scene recommendation list includes information on multiple smart scenes, and the information on the smart scenes is used to control the smart home devices in the smart home system to perform corresponding operations.

[0127] The second scene recommendation list acquisition module 602 is used to personalize the first scene recommendation list based on the local data of the smart home system to obtain the second scene recommendation list.

[0128] The intelligent scene recommendation module 603 is used to perform intelligent scene recommendation based on the second scene recommendation list.

[0129] In some embodiments of the present invention, the first scene recommendation list is personalized and adjusted according to the local data of the smart home system to obtain a second scene recommendation list, including: reordering multiple smart scenes included in the first scene recommendation list according to the local data of the smart home system to obtain the second scene recommendation list.

[0130] In some embodiments of the present invention, based on local data of the smart home system, multiple smart scenes included in the first scene recommendation list are reordered to obtain a second scene recommendation list, including:

[0131] When a change in local data of the smart home system is detected, information about changes in user intent or environment in the smart home system is determined.

[0132] Based on the user intent change information or environment change information, the multiple intelligent scenarios included in the first scenario recommendation list are reordered to obtain the second scenario recommendation list.

[0133] In some embodiments of the present invention, based on local data of the smart home system, multiple smart scenes included in the first scene recommendation list are reordered to obtain a second scene recommendation list, including:

[0134] Determine user negative feedback data from the local data of the smart home system;

[0135] Based on the user negative feedback data, the multiple intelligent scenes included in the first scene recommendation list are reordered to obtain the second scene recommendation list.

[0136] In some embodiments of the present invention, the user negative feedback data includes: relevant data recommended to the user but not selected by the user.

[0137] In some embodiments of the present invention, it further includes:

[0138] The difference item determination module is used to determine the difference items based on the local data and the cloud data of the cloud server;

[0139] The intelligent scene adjustment module is used to determine the target intelligent scene corresponding to the difference item from multiple intelligent scenes included in the first scene recommendation list, and adjust the information of the target intelligent scene.

[0140] In some embodiments of the present invention, it further includes:

[0141] The data copying module is used to copy the user's relevant data from the edge device of the smart home system where the user previously logged in when the system detects that the user has logged in for the first time.

[0142] In this embodiment of the invention, edge devices in a smart home system receive a first scene recommendation list sent by a cloud server. The first scene recommendation list includes information on multiple smart scenes, and the information on the smart scenes is used to control the smart home devices in the smart home system to perform corresponding operations. Then, based on the local data of the smart home system, the first scene recommendation list is personalized to obtain a second scene recommendation list. Based on the second scene recommendation list, smart scene recommendations are made. This realizes that the edge devices in the smart home system can make personalized adjustments based on the smart scenes recommended by the cloud server, thereby improving the effect of smart scene recommendations.

[0143] Some embodiments of the present invention also provide an electronic device, including a processor, a memory, and a computer program stored in the memory and capable of running on the processor, wherein the computer program, when executed by the processor, implements the method described above.

[0144] Some embodiments of the present invention also provide a computer-readable storage medium on which a computer program is stored, and which, when executed by a processor, implements the method described above.

[0145] Some embodiments of the present invention also provide a computer program product, including a computer program that, when executed by a processor, implements the method described above.

[0146] As the device embodiment is basically similar to the method embodiment, the description is relatively simple, and relevant parts can be found in the description of the method embodiment.

[0147] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, use and processing of the relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and corresponding operation portals are provided for users to choose to authorize or refuse.

[0148] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0149] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, apparatus, or computer program products. Therefore, embodiments of the present invention can take the form of entirely hardware embodiments, entirely software embodiments, or embodiments combining software and hardware aspects. Furthermore, embodiments of the present invention can take the form of computer program products implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0150] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, terminal devices (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing terminal device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing terminal device, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0151] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing terminal device to operate in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0152] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal equipment, causing a series of operational steps to be performed on the computer or other programmable terminal equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable terminal equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0153] Although preferred embodiments of the present invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the present invention.

[0154] 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 the aforementioned element.

[0155] The above provides a detailed description of the method and apparatus for scene recommendation in a smart home system. Specific examples have been used to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. At the same time, for those skilled in the art, 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 method for scene recommendation in a smart home system, characterized in that, The method includes: The system receives a first scene recommendation list sent by a cloud server; the first scene recommendation list includes information on multiple smart scenes, and the information on the smart scenes is used to control the smart home devices in the smart home system to perform corresponding operations. Based on the local data of the smart home system, the first scene recommendation list is personalized to obtain the second scene recommendation list; Based on the second scenario recommendation list, intelligent scenario recommendations are made.

2. The method according to claim 1, characterized in that, Based on the local data of the smart home system, the first scene recommendation list is personalized to obtain a second scene recommendation list, including: based on the local data of the smart home system, the multiple smart scenes included in the first scene recommendation list are reordered to obtain the second scene recommendation list.

3. The method according to claim 2, characterized in that, Based on the local data of the smart home system, the multiple smart scenes included in the first scene recommendation list are reordered to obtain the second scene recommendation list, which includes: When a change in local data of the smart home system is detected, information about changes in user intent or environment in the smart home system is determined. Based on the user intent change information or environment change information, the multiple intelligent scenarios included in the first scenario recommendation list are reordered to obtain the second scenario recommendation list.

4. The method according to claim 2, characterized in that, Based on the local data of the smart home system, the multiple smart scenes included in the first scene recommendation list are reordered to obtain the second scene recommendation list, which includes: Determine user negative feedback data from the local data of the smart home system; Based on the user negative feedback data, the multiple intelligent scenes included in the first scene recommendation list are reordered to obtain the second scene recommendation list.

5. The method according to claim 4, characterized in that, The negative user feedback data includes: relevant data that was recommended to the user but was not selected by the user.

6. The method according to any one of claims 2-5, characterized in that, Before reordering the multiple smart scenes included in the first scene recommendation list based on the local data of the smart home system to obtain the second scene recommendation list, the method further includes: Based on the local data and the cloud data from the cloud server, determine the differences; From the multiple intelligent scenarios included in the first scenario recommendation list, a target intelligent scenario corresponding to the difference item is determined, and the information of the target intelligent scenario is adjusted.

7. The method according to claim 1, characterized in that, The method is applied to edge devices and further includes: When a user logs into the smart home system for the first time, the user's relevant data is copied from the edge device of the smart home system in which the user previously logged in.

8. A device for scene recommendation in a smart home system, characterized in that, The device includes: The first scene recommendation list receiving module is used to receive a first scene recommendation list sent by the cloud server; the first scene recommendation list includes information on multiple smart scenes, and the information on the smart scenes is used to control the smart home devices in the smart home system to perform corresponding operations. The second scenario recommendation list acquisition module is used to personalize the first scenario recommendation list based on the local data of the smart home system to obtain the second scenario recommendation list; The intelligent scene recommendation module is used to make intelligent scene recommendations based on the second scene recommendation list.

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

10. A computer-readable storage medium, characterized in that, A computer program is stored on the computer-readable storage medium, which, when executed by a processor, implements the method as described in any one of claims 1 to 7.

11. A computer program product, characterized in that, It includes a computer program that, when executed by a processor, implements the method as described in any one of claims 1 to 7.

Images