Advertisement push method, system and device for smart toilets

Abstract

The advertising push method, system and device for smart toilets provided in the present application relate to the field of smart toilet technology. In the present application, firstly, for each time interval, a corresponding smart toilet user graph is constructed based on each smart toilet user who uses the smart toilet in the time interval and the user preference representation data corresponding to each smart toilet user; secondly, the smart toilet users matching between each two smart toilet user graphs in the multiple smart toilet user graphs corresponding to the multiple time intervals are connected and processed to form a total smart toilet user graph; then, for each time interval, based on the smart toilet user graph corresponding to the time interval and the total smart toilet user graph, at least one candidate advertisement is determined from multiple candidate advertisements as a target preference advertisement. Based on the above content, the problem of relatively low reliability of advertising push existing in the prior art can be improved.

Description

Technical Field

[0001] The present application relates to the technical field of smart toilets, and more specifically, to an advertising push method, system, and device for smart toilets. Background Art

[0002] With the continuous promotion of smart toilets, their number has increased dramatically. In the application of smart toilets, corresponding advertisements can generally be pushed. However, in the existing technology, fixed advertisements are generally pushed, or advertisements are pushed randomly in turn, or the length of the push time is determined based on the amount paid by the advertiser, and the time matching (or user matching) is not considered. Therefore, the existing advertisement push technology has the problem of relatively low reliability. Summary of the invention

[0003] In view of this, the purpose of the present application is to provide an advertising push method, system and device for smart toilets, so as to improve the problem of relatively low reliability of advertising push in the prior art.

[0004] To achieve the above objectives, this application adopts the following technical solutions:

[0005] A method for pushing advertisements for a smart toilet, comprising:

[0006] For each time interval, based on each smart toilet user who uses the smart toilet in the time interval and the user preference representation data corresponding to each smart toilet user, a smart toilet user graph corresponding to the time interval is constructed, wherein in the smart toilet user graph, any two connected smart toilet users have a correlation relationship based on the corresponding user preference representation data, and the user preference representation data is used to reflect the behavioral tendencies of the smart toilet users;

[0007] Connecting the smart toilet users that match between every two smart toilet user graphs in the multiple smart toilet user graphs corresponding to the multiple time intervals to form a corresponding smart toilet user total graph, wherein the matched smart toilet users refer to the same smart toilet users;

[0008] For each of the time intervals, based on the smart toilet user map corresponding to the time interval and the total smart toilet user map, at least one candidate advertisement matching the time interval is determined from multiple candidate advertisements as the corresponding target preference advertisement, wherein the advertising device in the smart toilet is used to play the corresponding target preference advertisement in the corresponding time interval.

[0009] In a preferred embodiment of the present application, in the above-mentioned method for pushing advertisements for smart toilets, for each of the time intervals, based on the smart toilet user graph corresponding to the time interval and the total graph of smart toilet users, the step of determining at least one candidate advertisement matching the time interval from a plurality of candidate advertisements as the corresponding target preference advertisement includes:

[0010] Performing feature mining on the user preference representation data corresponding to each smart toilet user in the overall map of smart toilet users to obtain the user preference features corresponding to each smart toilet user;

[0011] The user preference feature corresponding to each smart toilet user in the total smart toilet user map is transferred to obtain a first user preference transfer feature corresponding to each smart toilet user, and, for each time interval, the user preference feature corresponding to each smart toilet user in the smart toilet user map corresponding to the time interval is transferred to obtain a second user preference transfer feature corresponding to each smart toilet user;

[0012] For each of the time intervals, feature aggregation is performed on the first user preference transfer feature corresponding to each smart toilet user in the smart toilet user graph corresponding to the time interval to obtain the first preference aggregation feature corresponding to the time interval, and feature aggregation is performed on the second user preference transfer feature corresponding to each smart toilet user in the smart toilet user graph corresponding to the time interval to obtain the second preference aggregation feature corresponding to the time interval;

[0013] For each of the time intervals, based on the first preference aggregation feature and the second preference aggregation feature corresponding to the time interval, at least one candidate advertisement matching the time interval is determined from a plurality of candidate advertisements as the corresponding target preference advertisement.

[0014] In a preferred embodiment of the present application, in the above-mentioned method for pushing advertisements for smart toilets, the step of performing feature mining on the user preference representation data corresponding to each smart toilet user in the overall map of smart toilet users to obtain the user preference features corresponding to each smart toilet user includes:

[0015] Embedding the user preference representation data corresponding to each smart toilet user in the overall map of smart toilet users to obtain the preference embedding features corresponding to each smart toilet user;

[0016] The preference embedding features corresponding to each of the smart toilet users are processed by self-attention to obtain the corresponding preference attention features, and the preference embedding features corresponding to each of the smart toilets and the corresponding preference attention features are superimposed to obtain the user preference features corresponding to each of the smart toilet users.

[0017] In a preferred embodiment of the present application, in the above-mentioned method for pushing advertisements for smart toilets, the step of transferring the user preference features corresponding to each smart toilet user in the total map of smart toilet users to obtain the first user preference transfer features corresponding to each smart toilet user, and, for each of the time intervals, transferring the user preference features corresponding to each smart toilet user in the map of smart toilet users corresponding to the time interval to obtain the second user preference transfer features corresponding to each smart toilet user, includes:

[0018] Clustering the user preference features corresponding to each smart toilet user in the overall smart toilet user map to form at least one first central preference feature, and based on the at least one second central preference feature, performing feature transfer on the user preference features corresponding to each smart toilet user in the overall smart toilet user map to obtain a corresponding first user preference transfer feature;

[0019] For each of the time intervals, the user preference features corresponding to each smart toilet user in the smart toilet user graph corresponding to the time interval are clustered to form at least one second center user preference feature, and based on the at least one second center user preference feature, the user preference features corresponding to each smart toilet user in the smart toilet user graph corresponding to the time interval are feature transferred to obtain the corresponding second user preference transfer feature.

[0020] In a preferred embodiment of the present application, in the above-mentioned advertising push method for smart toilets, the step of clustering the user preference features corresponding to each smart toilet user in the total map of smart toilet users to form at least one first center preference feature, and based on the at least one second center preference feature, performing feature transfer on the user preference features corresponding to each smart toilet user in the total map of smart toilet users to obtain the corresponding first user preference transfer feature includes:

[0021] Counting the smart toilet users in the overall map of smart toilet users to obtain the number of user statistics corresponding to the overall map of smart toilet users, and determining the number of cluster centers based on the number of user statistics, wherein the number of cluster centers is less than the number of user statistics;

[0022] Based on the number of cluster centers, clustering the user preference features corresponding to each smart toilet user in the overall map of smart toilet users to form a corresponding number of first center preference features, and based on the number of user preference features in the cluster clusters corresponding to each of the first center preference features, determining the transmission priority corresponding to each of the first center preference features;

[0023] Based on the corresponding delivery priority, traverse each of the first center preference features in turn;

[0024] Taking the first central preference feature currently traversed as the starting point, the current feature transfer is performed on the user preference feature to be transferred corresponding to each smart toilet user in the total map of smart toilet users to form a transferred user preference feature, wherein the transferred user preference feature formed by the current feature transfer is used as the user preference feature to be transferred in the next feature transfer;

[0025] The transferred user preference feature formed by performing feature transfer starting from the last traversed first central preference feature is used as the corresponding first user preference transfer feature.

[0026] In a preferred embodiment of the present application, in the above-mentioned method for pushing advertisements for smart toilets, the step of taking the first central preference feature currently traversed as the starting point, performing the current feature transfer on the user preference feature to be transferred corresponding to each smart toilet user in the total map of smart toilet users, and forming the transferred user preference feature includes:

[0027] Performing self-attention processing on the first center preference feature currently traversed to obtain a corresponding first self-attention feature, and superimposing the first center preference feature and the first self-attention feature to obtain a transferred user preference feature;

[0028] For each other smart toilet user other than the smart toilet user corresponding to the first central preference feature in the overall graph of smart toilet users, determine the transfer level corresponding to the other smart toilet user based on the connectivity between the other smart toilet user and the smart toilet user corresponding to the first central preference feature in the overall graph of smart toilet users, wherein the connectivity is related to the path length of the shortest traversal path of the smart toilet user in the overall graph of smart toilet users;

[0029] Based on the corresponding transfer level, traverse each other smart toilet user in turn;

[0030] For each other smart toilet user currently traversed, based on the transferred user preference features of the smart toilet user of the previous transfer level connected to the other smart toilet user, the current feature transfer is performed on the user preference features to be transferred corresponding to the other smart toilet user to form the transferred user preference features.

[0031] In a preferred embodiment of the present application, in the above-mentioned method for pushing advertisements for smart toilets, the step of performing current feature transfer on the user preference features to be transferred corresponding to each other smart toilet user currently traversed, based on the transferred user preference features of the smart toilet user of the previous transfer level connected to the other smart toilet user, to form the transferred user preference features includes:

[0032] For each other smart toilet user currently traversed, the association parameters between the transferred user preference features of each smart toilet user at the previous transfer level connected to the other smart toilet user and the user preference features to be transferred corresponding to the other smart toilet user are determined respectively, and based on the association parameters, the user preference features to be transferred corresponding to the other smart toilet user are adjusted, and the corresponding adjusted user preference features are output;

[0033] For each other smart toilet user currently traversed, the mean of each adjusted user preference feature corresponding to the other smart toilet user is calculated to obtain the corresponding mean user preference feature, and the mean user preference feature and the user preference feature to be transmitted corresponding to the other smart toilet user are superimposed to obtain the transmitted user preference feature.

[0034] In a preferred embodiment of the present application, in the above-mentioned method for pushing advertisements for smart toilets, for each of the time intervals, based on the first preference aggregation feature and the corresponding second preference aggregation feature corresponding to the time interval, at least one candidate advertisement matching the time interval is determined from multiple candidate advertisements as the corresponding target preference advertisement, including:

[0035] For each of the time intervals, feature aggregation is performed on the user preference features corresponding to each smart toilet user in the smart toilet user graph corresponding to the time interval to obtain a third preference aggregation feature corresponding to the time interval, and the first preference aggregation feature, the second preference aggregation feature and the third preference aggregation feature corresponding to the time interval are averaged to obtain a target preference aggregation feature corresponding to the time interval;

[0036] For each candidate advertisement among the multiple candidate advertisements, feature mining is performed on the candidate advertisement to obtain advertisement content features corresponding to the candidate advertisement, and feature correlations between the advertisement content features and target preference aggregation features corresponding to each time interval are determined respectively to obtain push matching parameters between the candidate advertisement and each time interval;

[0037] For each of the time intervals, based on the push matching parameters between the time interval and each of the candidate advertisements, at least one candidate advertisement matching the time interval is determined from among the multiple candidate advertisements as the corresponding target preference advertisement.

[0038] This application also provides an advertising push system for smart toilets, including:

[0039] A toilet user graph construction module is used to construct a smart toilet user graph corresponding to each time interval based on each smart toilet user who uses the smart toilet in the time interval and the user preference representation data corresponding to each smart toilet user, wherein in the smart toilet user graph, any two connected smart toilet users have a correlation relationship based on the corresponding user preference representation data, and the user preference representation data is used to reflect the behavioral tendencies of the smart toilet users;

[0040] A toilet user graph connection module is used to connect the smart toilet users that match between every two smart toilet user graphs in a plurality of smart toilet user graphs corresponding to a plurality of time intervals to form a corresponding smart toilet user total graph, wherein the matched smart toilet users refer to the same smart toilet users;

[0041] The target preference advertisement determination module is used to determine, for each of the time intervals, at least one candidate advertisement matching the time interval from a plurality of candidate advertisements based on the smart toilet user map corresponding to the time interval and the total smart toilet user map, as the corresponding target preference advertisement, wherein the advertising device in the smart toilet is used to play the corresponding target preference advertisement in the corresponding time interval.

[0042] Based on the above, the present application also provides an electronic device, including:

[0043] Memory for storing computer programs;

[0044] A processor connected to the memory is used to execute the computer program stored in the memory to implement the above-mentioned advertising push method for smart toilets.

[0045] The advertising push method, system and device for smart toilets provided in the present application firstly, for each time interval, based on each smart toilet user who uses the smart toilet in the time interval and the user preference representation data corresponding to each smart toilet user, a corresponding smart toilet user graph is constructed; secondly, the smart toilet users matching between each two smart toilet user graphs in the multiple smart toilet user graphs corresponding to the multiple time intervals are connected to form a total smart toilet user graph; then, for each time interval, based on the smart toilet user graph corresponding to the time interval and the total smart toilet user graph, at least one candidate advertisement is determined from multiple candidate advertisements as a target preference advertisement. Based on the above content, since the graph is first constructed in time intervals, and then the graphs are connected based on the smart toilet users, the total smart toilet user graph formed has both high accuracy and rich semantic information, so that the target preference advertisement can be reliably determined for the corresponding time interval based on the total smart toilet user graph and the corresponding smart toilet user graph, that is, the reliability of advertisement push is guaranteed, thereby improving the problem of relatively low reliability of advertisement push in the prior art. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, preferred embodiments are specifically cited below and described in detail with reference to the attached drawings.

[0047] Figure 1 A structural block diagram of an electronic device provided in an embodiment of the present application.

[0048] Figure 2 A flowchart of an advertising push method for a smart toilet provided in an embodiment of the present application.

[0049] Figure 3 A schematic diagram of the overall map of smart toilet users provided in an embodiment of the present application.

[0050] Figure 4 A block diagram of an advertising push system for a smart toilet provided in an embodiment of the present application. DETAILED DESCRIPTION

[0051] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. The components of the embodiments of the present application described and shown in the drawings here can be arranged and designed in various different configurations.

[0052] Therefore, the following detailed description of the embodiments of the present application provided in the accompanying drawings is not intended to limit the scope of the present application for which protection is sought, but merely represents selected embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work are within the scope of protection of the present application.

[0053] like Figure 1 As shown, an embodiment of the present application provides an electronic device, wherein the electronic device may include a memory, a processor, and an advertisement push system for a smart toilet.

[0054] In detail, the memory and the processor are electrically connected directly or indirectly to achieve data transmission or interaction. For example, the memory and the processor can be electrically connected through one or more communication buses or signal lines. The advertising push system for smart toilets includes at least one software function module stored in the memory in the form of software or firmware. The processor is used to execute an executable computer program stored in the memory, for example, the software function modules and computer programs included in the advertising push system for smart toilets, so as to implement the advertising push method for smart toilets provided in the embodiment of the present application.

[0055] Optionally, the memory may be, but is not limited to, a random access memory (RAM), a read only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable read-only memory (EEPROM), etc.

[0056] Furthermore, the processor may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), a system on chip (SoC), etc.; it may also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0057] Understandably, Figure 1The structure shown is for illustration only, and the electronic device may also include Figure 1 More or fewer components as shown, or with Figure 1 The different configurations shown, for example, may also include a communication unit for exchanging information with other devices (such as advertising devices).

[0058] Combination Figure 2 The embodiment of the present application also provides an advertisement push method for smart toilets that can be applied to the above electronic device. The method steps defined in the process related to the advertisement push method for smart toilets can be implemented by the electronic device.

[0059] The following will Figure 2 The specific process shown is explained in detail.

[0060] Step S110, for each time interval, based on each smart toilet user who uses the smart toilet in the time interval and the user preference representation data corresponding to each smart toilet user, construct a smart toilet user map corresponding to the time interval.

[0061] In the embodiment of the present application, the electronic device can construct a smart toilet user map corresponding to each time interval (for example, a day is divided into multiple time intervals according to a preset time interval, such as 8 to 9 o'clock, 9 to 10 o'clock, 10 to 11 o'clock, 11 to 12 o'clock, 12 to 13 o'clock, 13 to 14 o'clock, etc.) based on the user preference characterization data corresponding to each smart toilet user who uses the smart toilet in the time interval. Figure 3 In the smart toilet user graph, any two connected smart toilet users have a correlation based on the corresponding user preference representation data. The user preference representation data is used to reflect the behavioral tendencies of the smart toilet users. Exemplarily, the user preference representation data may include the registration information and usage information of the smart toilet users on the corresponding smart toilet platform, such as age, gender, occupation, hobbies and other information related to user preferences that can be collected, and historical usage information on the smart toilet platform, including browsing and search information. In addition, the user preference representation data having a correlation may refer to the same or similar information such as age, or the edit distance between the user preference representation data may be directly calculated, and then the user preference representation data with an edit distance less than a preset distance may be determined to have a correlation.

[0062] Step S120, connecting the smart toilet users that match between every two smart toilet user graphs in the multiple smart toilet user graphs corresponding to the multiple time intervals to form a corresponding smart toilet user total graph.

[0063] In an embodiment of the present application, the electronic device can connect the smart toilet users that match between each two smart toilet user graphs in the multiple smart toilet user graphs corresponding to the multiple time intervals to form a corresponding smart toilet user total graph. The matched smart toilet users refer to the same smart toilet users, that is, the same smart toilet user, such as the same registration information.

[0064] Step S130, for each of the time intervals, based on the smart toilet user graph corresponding to the time interval and the total smart toilet user graph, at least one candidate advertisement matching the time interval is determined from multiple candidate advertisements as the corresponding target preference advertisement.

[0065] In an embodiment of the present application, the electronic device can determine, for each time interval, at least one candidate advertisement matching the time interval from among multiple candidate advertisements based on the smart toilet user graph corresponding to the time interval and the overall smart toilet user graph, as the corresponding target preference advertisement. The advertising device (a device capable of displaying video or images) in the smart toilet is used to play the corresponding target preference advertisement in the corresponding time interval.

[0066] Based on the above content, since the graph is first constructed in time intervals and then the graphs are connected based on smart toilet users, the overall graph of smart toilet users formed has both high accuracy and rich semantic information, so that the target preference advertisements for the corresponding time intervals can be reliably determined based on the overall graph of smart toilet users and the corresponding smart toilet user graph, that is, the reliability of advertisement push is guaranteed, thereby improving the problem of relatively low reliability of advertisement push in the prior art.

[0067] It should be noted that for step S130 , the specific method of determining at least one candidate advertisement matching the time interval from a plurality of candidate advertisements is not limited and can be selected according to actual needs.

[0068] For example, in an alternative implementation, in order to improve the reliability of the determined matching candidate advertisements, the above-mentioned step S130 may further include step S131, step S132, step S133 and step S134, and the specific content of each step is described as follows.

[0069] Step S131, performing feature mining on the user preference representation data corresponding to each smart toilet user in the overall map of smart toilet users to obtain the user preference features corresponding to each smart toilet user.

[0070] In an embodiment of the present application, feature mining can be performed on the user preference representation data corresponding to each smart toilet user in the overall map of smart toilet users to obtain user preference features corresponding to each smart toilet user. For example, feature mining can be performed on the user preference representation data corresponding to smart toilet user 1 to obtain the user preference features corresponding to smart toilet user 1; feature mining can be performed on the user preference representation data corresponding to smart toilet user 2 to obtain the user preference features corresponding to smart toilet user 2; feature mining can be performed on the user preference representation data corresponding to smart toilet user 3 to obtain the user preference features corresponding to smart toilet user 3.

[0071] Step S132, feature transfer is performed on the user preference features corresponding to each smart toilet user in the total smart toilet user map to obtain the first user preference transfer features corresponding to each smart toilet user, and, for each of the time intervals, feature transfer is performed on the user preference features corresponding to each smart toilet user in the smart toilet user map corresponding to the time interval to obtain the second user preference transfer features corresponding to each smart toilet user.

[0072] In an embodiment of the present application, after obtaining the user preference features corresponding to each smart toilet user, the user preference features corresponding to each smart toilet user in the total map of smart toilet users can be feature transferred to obtain the first user preference transfer features corresponding to each smart toilet user, and, for each time interval, the user preference features corresponding to each smart toilet user in the smart toilet user map corresponding to the time interval can be feature transferred to obtain the second user preference transfer features corresponding to each smart toilet user. In other words, on the one hand, feature transfer can be performed inside the total map of smart toilet users, so that the semantic information in the total map of smart toilet users can be fully integrated to form a semantically rich first user preference transfer feature. On the other hand, feature transfer can be performed separately inside each smart toilet user map, so that the semantic information inside each smart toilet user map can also be fully integrated to form a semantically rich second user preference transfer feature. Based on this, semantic fusion of different granularities can be achieved, so that when enriching semantic information, accuracy can also be taken into account.

[0073] Step S133, for each of the time intervals, feature aggregation is performed on the first user preference transfer features corresponding to each smart toilet user in the smart toilet user graph corresponding to the time interval to obtain the first preference aggregation features corresponding to the time interval, and feature aggregation is performed on the second user preference transfer features corresponding to each smart toilet user in the smart toilet user graph corresponding to the time interval to obtain the second preference aggregation features corresponding to the time interval.

[0074] In an embodiment of the present application, for each of the time intervals, the first user preference transfer features corresponding to each smart toilet user in the smart toilet user graph corresponding to the time interval are feature aggregated (such as the concatenation or mean calculation of the first user preference transfer features to achieve aggregation) to obtain the first preference aggregation feature corresponding to the time interval, and the second user preference transfer features corresponding to each smart toilet user in the smart toilet user graph corresponding to the time interval are feature aggregated to obtain the second preference aggregation feature corresponding to the time interval (such as the concatenation or mean calculation of the second user preference transfer features to achieve aggregation).

[0075] Step S134, for each of the time intervals, based on the first preference aggregation feature and the second preference aggregation feature corresponding to the time interval, at least one candidate advertisement matching the time interval is determined from multiple candidate advertisements as the corresponding target preference advertisement.

[0076] In an embodiment of the present application, after obtaining the semantically rich first preference aggregation features and second preference aggregation features, for each of the time intervals, based on the first preference aggregation features and the corresponding second preference aggregation features corresponding to the time interval, at least one candidate advertisement matching the time interval can be determined from multiple candidate advertisements as the corresponding target preference advertisement.

[0077] It can be understood that in the above step S131, the specific method of performing feature mining on the user preference representation data corresponding to each smart toilet user in the overall map of smart toilet users is not limited and can be selected according to actual needs. For example, in an alternative implementation, in order to improve the semantic representation ability of the mined user preference features, the above step S131 can further include the following content:

[0078] First, the user preference representation data corresponding to each smart toilet user in the overall smart toilet user graph may be embedded to obtain the preference embedding features corresponding to each smart toilet user; illustratively, the user preference representation data may be embedded using a trained word embedding model to obtain the corresponding preference embedding features;

[0079] Secondly, the preference embedding features corresponding to each of the smart toilet users can be processed by self-attention (the specific processing can refer to the relevant existing technology and will not be described here one by one) to obtain the preference attention features corresponding to each of the smart toilet users, and the preference embedding features corresponding to each of the smart toilets and the corresponding preference attention features can be superimposed to obtain the user preference features corresponding to each smart toilet user. In this way, the user preference features can focus on important features and avoid losing some details in the self-attention processing, that is, to ensure that the obtained user preference features have a more reliable semantic representation ability.

[0080] It can be understood that in the above-mentioned step S132, the specific method of transmitting the user preference characteristics corresponding to each smart toilet user is not restricted and can be selected according to actual needs. For example, in an alternative embodiment, in order to ensure the reliability of feature transmission, the above-mentioned step S132 can further include step S132a and step S132b, the specific contents of which are as follows.

[0081] Step S132a, clustering the user preference features corresponding to each smart toilet user in the total smart toilet user map to form at least one first center preference feature, and based on the at least one first center preference feature, performing feature transfer on the user preference features corresponding to each smart toilet user in the total smart toilet user map to obtain the corresponding first user preference transfer feature.

[0082] In the embodiment of the present application, the user preference features corresponding to each smart toilet user in the overall map of smart toilet users can be clustered to form at least one first center preference feature, and based on the at least one first center preference feature, the user preference features corresponding to each smart toilet user in the overall map of smart toilet users can be feature transferred to obtain the corresponding first user preference transfer feature. In this way, feature transfer based on cluster centers can not only ensure the reliability of the transferred features, but also reduce the complexity of feature transfer to a certain extent.

[0083] Step S132b, for each of the time intervals, clustering the user preference features corresponding to each smart toilet user in the smart toilet user graph corresponding to the time interval to form at least one second center user preference feature, and based on the at least one second center user preference feature, feature transfer is performed on the user preference features corresponding to each smart toilet user in the smart toilet user graph corresponding to the time interval to obtain the corresponding second user preference transfer feature.

[0084] In an embodiment of the present application, for each of the time intervals, the user preference features corresponding to each smart toilet user in the smart toilet user graph corresponding to the time interval can be clustered (which can be consistent with the clustering processing method in the previous step) to form at least one second center user preference feature, and based on the at least one second center user preference feature, the user preference features corresponding to each smart toilet user in the smart toilet user graph corresponding to the time interval can be feature transferred (which can be consistent with the feature transfer method in the previous step) to obtain the corresponding second user preference transfer feature.

[0085] It can be understood that, in the above-mentioned step S132a, the specific method of clustering user preference features and transferring features based on the obtained first center preference features is not limited and can be selected according to actual needs. For example, in an alternative embodiment, in order to ensure the reliability of the obtained first user preference transfer features, the above-mentioned step S132a can further include step a1, step a2, step a3, step a4 and step a5, the specific contents of which are as follows.

[0086] Step a1, counting the smart toilet users in the total map of smart toilet users, obtaining the statistical number of users corresponding to the total map of smart toilet users, and determining the number of cluster centers based on the statistical number of users.

[0087] In an embodiment of the present application, the smart toilet users in the total map of smart toilet users can be counted to obtain the number of user statistics corresponding to the total map of smart toilet users, and the number of cluster centers can be determined based on the number of user statistics. The number of cluster centers is less than the number of user statistics. For example, a reference ratio can be pre-configured, such as 0.05, 0.1, 0.15, 0.2, etc., and then the product of the number of user statistics and the reference ratio is used as the number of cluster centers.

[0088] Step a2, based on the number of clustering centers, cluster the user preference characteristics corresponding to each smart toilet user in the total smart toilet user map to form a corresponding number of first center preference characteristics, and based on the number of user preference characteristics in the cluster clusters corresponding to each of the first center preference characteristics, determine the transmission priority corresponding to each of the first center preference characteristics.

[0089] In the embodiment of the present application, after obtaining the number of cluster centers, the user preference features corresponding to each smart toilet user in the total map of smart toilet users can be clustered based on the number of cluster centers to form a corresponding number of first center preference features, and the transmission priority corresponding to each first center preference feature can be determined based on the number of user preference features in the cluster cluster corresponding to each first center preference feature. Exemplarily, a corresponding number of user preference features can be randomly determined as initial center preference features based on the number of cluster centers, and then, for each user preference feature, the user preference feature is assigned to the cluster cluster corresponding to the nearest initial center preference feature, and for each cluster cluster, the feature distance (such as cosine distance, etc.) between each user preference feature in the cluster cluster and the initial center preference feature can be calculated, and the mean of each feature distance can be calculated to obtain the distance representation value corresponding to the cluster cluster, and the distance representation value corresponding to each cluster cluster is weighted averaged (the weighting coefficient has a positive correlation with the number of user preference features included in the cluster cluster) to obtain the corresponding target distance representation value. Based on this, multiple target distance representation values ​​can be obtained through multiple randomizations, and finally, the initial center preference feature corresponding to the minimum target distance representation value can be selected as the first center preference feature. In addition, there can be a positive correlation between the number of user preference features in the cluster and the transfer priority, that is, the larger the number, the higher the transfer priority. In addition, when the number is the same, the transfer priority can be determined arbitrarily.

[0090] Step a3: traverse each of the first center preference features in turn based on the corresponding transfer priority.

[0091] In the embodiment of the present application, after the corresponding delivery priority is determined, each of the first center preference features can be traversed in sequence based on the corresponding delivery priority. For example, the first first center preference feature traversed has the highest delivery priority, and the last first center preference feature traversed has the lowest delivery priority.

[0092] Step a4, taking the first central preference feature currently traversed as the starting point, the current feature transfer is performed on the user preference feature to be transferred corresponding to each smart toilet user in the total map of smart toilet users to form the transferred user preference feature.

[0093] In an embodiment of the present application, after determining the first center preference feature currently traversed, the first center preference feature currently traversed can be used as a starting point to perform the current feature transfer for the user preference feature to be transferred corresponding to each smart toilet user in the overall map of smart toilet users, to form the transferred user preference feature, that is, the semantic information of the first center preference feature currently traversed verifies the links in the overall map of smart toilet users and transfers them in sequence, so that other user preference features can be updated, thereby forming the transferred user preference feature. Among them, the transferred user preference feature formed by the current feature transfer is used as the user preference feature to be transferred for the next feature transfer, so that the second first center preference feature traversed and each first center preference feature traversed thereafter are actually the user preference features after the previous feature transfer, that is, the user preference features to be transferred.

[0094] Step a5, taking the last traversed first central preference feature as the starting point for feature transfer to form a transferred user preference feature as the corresponding first user preference transfer feature.

[0095] In an embodiment of the present application, the user preference feature after feature transfer formed by taking the last traversed first center preference feature as the starting point can be used as the corresponding first user preference transfer feature, that is, the user preference feature after feature transfer formed by the last feature transfer.

[0096] It is understandable that, in the above step a4, the specific manner of performing the current feature transfer on the user preference feature to be transferred is not limited and can be selected according to actual needs. For example, in an alternative implementation, the above step a4 may include the following contents:

[0097] First, the first center preference feature currently traversed (as mentioned above, the second first center preference feature traversed and each first center preference feature traversed thereafter are actually the user preference features after the last feature transfer, for example, the second first center preference feature traversed is actually the user preference feature after feature transfer based on the first first center preference feature traversed) is processed by self-attention to obtain the corresponding first self-attention feature, and the first center preference feature is superimposed with the first self-attention feature to obtain the transferred user preference feature;

[0098] Secondly, for each other smart toilet user other than the smart toilet user corresponding to the first central preference feature in the total graph of smart toilet users, the transmission level corresponding to the other smart toilet user is determined based on the connectivity between the other smart toilet user and the smart toilet user corresponding to the first central preference feature in the total graph of smart toilet users, wherein the connectivity is related to the path length of the shortest traversal path of the smart toilet user in the total graph of smart toilet users, such as a negative correlation, that is, the shorter the path length, the greater the connectivity, and the higher the corresponding transmission level, wherein for the shortest traversal path, such as Figure 3 For smart toilet user 2 and smart toilet user 5, the traversal paths can be "smart toilet user 2-smart toilet user 5" and "smart toilet user 2-smart toilet user 3-smart toilet user 5", among which "smart toilet user 2-smart toilet user 5" is the shortest traversal path;

[0099] Then, based on the corresponding transfer level, traverse each other smart toilet user in turn. For example, the first other smart toilet user traversed has the highest corresponding transfer level, and the last other smart toilet user traversed has the lowest corresponding transfer level.

[0100] Finally, for each other smart toilet user currently traversed, based on the transferred user preference features of the smart toilet user of the previous transfer level connected to the other smart toilet user, the current feature transfer is performed on the user preference features to be transferred corresponding to the other smart toilet user to form the transferred user preference features. For example, in the first step, the transferred user preference features corresponding to smart toilet user 2 can be transferred to the user preference features to be transferred corresponding to smart toilet user 3, smart toilet user 4 and smart toilet user 5, respectively, to form the transferred user preference features corresponding to smart toilet user 3, smart toilet user 4 and smart toilet user 5, respectively. In the second step, the transferred user preference features corresponding to smart toilet user 3 can be transferred to the user preference features to be transferred corresponding to smart toilet user 3 in another smart toilet user graph, and the transferred user preference features corresponding to smart toilet user 4 can be transferred to the user preference features to be transferred corresponding to smart toilet user 1, and the transferred user preference features corresponding to smart toilet user 5 can be transferred to the user preference features to be transferred corresponding to smart toilet user 7 and smart toilet user 8, respectively. In this way, the transfer can be performed in sequence based on the transfer level.

[0101] It can be understood that in the above content, the specific method of performing the current feature transfer for the user preference features to be transferred corresponding to other smart toilet users is not limited and can be selected according to actual needs. For example, in an alternative implementation, in order to improve the semantic representation accuracy of the user preference features after transfer, the current feature transfer can be performed based on the following content:

[0102] First, for each other smart toilet user currently traversed, determine the association parameters (such as the result of multiplication) between the transferred user preference features of each smart toilet user at the previous transfer level connected to the other smart toilet user and the user preference features to be transferred corresponding to the other smart toilet user, and adjust the user preference features to be transferred corresponding to the other smart toilet user based on the association parameters (such as multiplying the association parameters with the user preference features to be transferred), and output the corresponding adjusted user preference features;

[0103] Secondly, for each other smart toilet user currently traversed, the mean of each adjusted user preference feature corresponding to the other smart toilet user is calculated to obtain the corresponding mean user preference feature, and the mean user preference feature and the user preference feature to be transmitted corresponding to the other smart toilet user are superimposed to obtain the transmitted user preference feature.

[0104] It can be understood that, in the above step S134, the specific method of determining at least one candidate advertisement matching the time interval from multiple candidate advertisements based on the first preference aggregation feature corresponding to the time interval and the corresponding second preference aggregation feature is not limited and can be selected according to actual needs. For example, in an alternative implementation, in order to further improve the reliability of the determined matching candidate advertisements, the above step S134 can further include the following contents:

[0105] First, for each of the time intervals, the user preference features corresponding to each smart toilet user in the smart toilet user graph corresponding to the time interval are aggregated (as described above) to obtain the third preference aggregation feature corresponding to the time interval, and the first preference aggregation feature, the second preference aggregation feature and the third preference aggregation feature corresponding to the time interval are averaged to obtain the target preference aggregation feature corresponding to the time interval;

[0106] Secondly, for each candidate advertisement among the multiple candidate advertisements, feature mining is performed on the candidate advertisement (for example, a pre-trained convolutional neural network can be used to perform convolution processing on the video frame or image frame corresponding to the candidate advertisement), and the advertisement content feature corresponding to the candidate advertisement is obtained, and the feature correlation between the advertisement content feature and the target preference aggregation feature corresponding to each time interval is determined respectively, and the push matching parameter between the candidate advertisement and each time interval is obtained. For example, the cosine similarity between the features can be used as the push matching parameter;

[0107] Then, for each of the time intervals, based on the push matching parameters between the time interval and each of the candidate advertisements, at least one candidate advertisement matching the time interval is determined among the multiple candidate advertisements as the corresponding target preference advertisement. For example, one or more candidate advertisements with the largest push matching parameters can be screened out as the target preference advertisements.

[0108] Combination Figure 4 The embodiment of the present application also provides an advertisement push system for smart toilets that can be applied to the above electronic device. The advertisement push system for smart toilets can include a toilet user graph construction module, a toilet user graph connection module, and a target preference advertisement determination module.

[0109] The toilet user map construction module is used to construct a smart toilet user map corresponding to each time interval based on the user preference characterization data corresponding to each smart toilet user who uses the smart toilet in the time interval, wherein in the smart toilet user map, any two connected smart toilet users have a correlation relationship based on the corresponding user preference characterization data, and the user preference characterization data is used to reflect the behavioral tendencies of the smart toilet users. In the embodiment of the present application, the toilet user map construction module can be used to execute Figure 2 As shown in step S110, the relevant contents of the toilet user map construction module can refer to the above description of step S110.

[0110] The toilet user map connection module is used to connect the smart toilet users that match between each two smart toilet user maps in the multiple smart toilet user maps corresponding to the multiple time intervals to form a corresponding smart toilet user overall map, wherein the matched smart toilet users refer to the same smart toilet users. Figure 2 As shown in step S120, the relevant contents of the toilet user map connection module can refer to the description of step S120 in the previous text.

[0111] The target preference advertisement determination module is used to determine, for each of the time intervals, at least one candidate advertisement matching the time interval from among multiple candidate advertisements based on the smart toilet user graph corresponding to the time interval and the total smart toilet user graph, as the corresponding target preference advertisement, wherein the advertising device in the smart toilet is used to play the corresponding target preference advertisement in the corresponding time interval. In the embodiment of the present application, the target preference advertisement determination module can be used to execute Figure 2 As shown in step S130, the relevant contents of the target preference advertisement determination module can refer to the above description of step S130.

[0112] In an embodiment of the present application, corresponding to the above-mentioned advertising push method for smart toilets applied to the electronic device, a computer-readable storage medium is also provided, in which a computer program is stored, and when the computer program is run, each step of the advertising push method for smart toilets is executed.

[0113] Among them, the steps executed when the aforementioned computer program is running will not be described one by one here, and reference can be made to the previous explanation of the advertising push method for smart toilets.

[0114] In summary, the advertising push method, system and device for smart toilets provided by the present application, first, for each time interval, based on each smart toilet user who uses the smart toilet in the time interval and the user preference representation data corresponding to each smart toilet user, a corresponding smart toilet user graph is constructed; secondly, the smart toilet users matching between each two smart toilet user graphs in the multiple smart toilet user graphs corresponding to the multiple time intervals are connected to form a total smart toilet user graph; then, for each time interval, based on the smart toilet user graph corresponding to the time interval and the total smart toilet user graph, at least one candidate advertisement is determined from multiple candidate advertisements as a target preference advertisement. Based on the above content, since the graph is first constructed in time intervals, and then the graphs are connected based on the smart toilet users, the total smart toilet user graph formed has both high accuracy and rich semantic information, so that the target preference advertisement can be reliably determined for the corresponding time interval based on the total smart toilet user graph and the corresponding smart toilet user graph, that is, the reliability of advertising push is guaranteed, thereby improving the problem of relatively low reliability of advertising push in the prior art.

[0115] In several embodiments provided in the embodiments of the present application, it should be understood that the disclosed device and method can also be implemented in other ways. The device and method embodiments described above are merely schematic, for example, the flowcharts and block diagrams in the accompanying drawings show the possible architecture, functions and operations of the device, method and computer program product according to multiple embodiments of the present application. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of a code, and the module, a program segment or a part of a code contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flowchart, and the combination of boxes in the block diagram and / or flowchart can be implemented with a dedicated hardware-based system that performs a specified function or action, or can be implemented with a combination of dedicated hardware and computer instructions.

[0116] In addition, the functional modules in the various embodiments of the present application may be integrated together to form an independent part, or each module may exist separately, or two or more modules may be integrated to form an independent part.

[0117] If the function is implemented in the form of a software function module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or the part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, electronic device, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program code. It should be noted that in this article, the term "include", "include" or any other variant thereof is intended to cover non-exclusive inclusion, so that the process, method, article or device including a series of elements includes not only those elements, but also includes other elements that are not explicitly listed, or also includes elements inherent to such process, method, article or device. Without more constraints, an element defined by the phrase "comprising a..." does not exclude the existence of other identical elements in the process, method, article or apparatus comprising the element.

[0118] The above description is only the preferred embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for pushing advertisements for smart toilets, characterized in that: include: For each time interval, based on each smart toilet user who uses the smart toilet in the time interval and the user preference representation data corresponding to each smart toilet user, a smart toilet user graph corresponding to the time interval is constructed, wherein in the smart toilet user graph, any two connected smart toilet users have a correlation relationship based on the corresponding user preference representation data, and the user preference representation data is used to reflect the behavioral tendencies of the smart toilet users; Connecting the smart toilet users that match between every two smart toilet user graphs in the multiple smart toilet user graphs corresponding to the multiple time intervals to form a corresponding smart toilet user total graph, wherein the matched smart toilet users refer to the same smart toilet users; For each of the time intervals, based on the smart toilet user graph corresponding to the time interval and the overall smart toilet user graph, at least one candidate advertisement matching the time interval is determined from a plurality of candidate advertisements as the corresponding target preference advertisement, wherein the advertising device in the smart toilet is used to play the corresponding target preference advertisement in the corresponding time interval; The step of determining, for each of the time intervals, based on the smart toilet user graph corresponding to the time interval and the total smart toilet user graph, at least one candidate advertisement matching the time interval from a plurality of candidate advertisements as the corresponding target preference advertisement comprises: Performing feature mining on the user preference representation data corresponding to each smart toilet user in the overall map of smart toilet users to obtain the user preference features corresponding to each smart toilet user; The user preference feature corresponding to each smart toilet user in the total smart toilet user map is transferred to obtain a first user preference transfer feature corresponding to each smart toilet user, and, for each time interval, the user preference feature corresponding to each smart toilet user in the smart toilet user map corresponding to the time interval is transferred to obtain a second user preference transfer feature corresponding to each smart toilet user; For each of the time intervals, feature aggregation is performed on the first user preference transfer feature corresponding to each smart toilet user in the smart toilet user graph corresponding to the time interval to obtain the first preference aggregation feature corresponding to the time interval, and feature aggregation is performed on the second user preference transfer feature corresponding to each smart toilet user in the smart toilet user graph corresponding to the time interval to obtain the second preference aggregation feature corresponding to the time interval; For each of the time intervals, based on the first preference aggregation feature and the second preference aggregation feature corresponding to the time interval, at least one candidate advertisement matching the time interval is determined from a plurality of candidate advertisements as the corresponding target preference advertisement; The step of transferring the user preference features corresponding to each smart toilet user in the overall smart toilet user map to obtain the first user preference transfer features corresponding to each smart toilet user, and, for each time interval, transferring the user preference features corresponding to each smart toilet user in the smart toilet user map corresponding to the time interval to obtain the second user preference transfer features corresponding to each smart toilet user, includes: Clustering the user preference features corresponding to each smart toilet user in the overall smart toilet user map to form at least one first central preference feature, and based on the at least one first central preference feature, performing feature transfer on the user preference features corresponding to each smart toilet user in the overall smart toilet user map to obtain a corresponding first user preference transfer feature; For each of the time intervals, clustering is performed on the user preference features corresponding to each smart toilet user in the smart toilet user graph corresponding to the time interval to form at least one second central user preference feature, and based on the at least one second central user preference feature, feature transfer is performed on the user preference features corresponding to each smart toilet user in the smart toilet user graph corresponding to the time interval to obtain a corresponding second user preference transfer feature; The step of clustering the user preference features corresponding to each smart toilet user in the overall smart toilet user map to form at least one first center preference feature, and based on the at least one first center preference feature, performing feature transfer on the user preference features corresponding to each smart toilet user in the overall smart toilet user map to obtain the corresponding first user preference transfer feature includes: Counting the smart toilet users in the overall map of smart toilet users to obtain the number of user statistics corresponding to the overall map of smart toilet users, and determining the number of cluster centers based on the number of user statistics, wherein the number of cluster centers is less than the number of user statistics; Based on the number of cluster centers, clustering the user preference features corresponding to each smart toilet user in the overall map of smart toilet users to form a corresponding number of first center preference features, and based on the number of user preference features in the cluster clusters corresponding to each of the first center preference features, determining the transmission priority corresponding to each of the first center preference features; Based on the corresponding delivery priority, traverse each of the first center preference features in turn; Taking the first central preference feature currently traversed as the starting point, the current feature transfer is performed on the user preference feature to be transferred corresponding to each smart toilet user in the total map of smart toilet users to form a transferred user preference feature, wherein the transferred user preference feature formed by the current feature transfer is used as the user preference feature to be transferred in the next feature transfer; The transferred user preference feature formed by performing feature transfer starting from the last traversed first central preference feature is used as the corresponding first user preference transfer feature.

2. The method for pushing advertisements for smart toilets according to claim 1, characterized in that: The step of performing feature mining on the user preference representation data corresponding to each smart toilet user in the overall map of smart toilet users to obtain the user preference features corresponding to each smart toilet user includes: Embedding the user preference representation data corresponding to each smart toilet user in the overall map of smart toilet users to obtain the preference embedding features corresponding to each smart toilet user; The preference embedding features corresponding to each of the smart toilet users are processed by self-attention to obtain the corresponding preference attention features, and the preference embedding features corresponding to each of the smart toilets and the corresponding preference attention features are superimposed to obtain the user preference features corresponding to each of the smart toilet users.

3. The advertising push method for smart toilets according to claim 1, characterized in that: The step of taking the first central preference feature currently traversed as a starting point, performing current feature transfer on the user preference feature to be transferred corresponding to each smart toilet user in the overall map of smart toilet users, and forming the transferred user preference feature, comprises: Performing self-attention processing on the first center preference feature currently traversed to obtain a corresponding first self-attention feature, and superimposing the first center preference feature and the first self-attention feature to obtain a transferred user preference feature; For each other smart toilet user other than the smart toilet user corresponding to the first central preference feature in the overall graph of smart toilet users, determine the transfer level corresponding to the other smart toilet user based on the connectivity between the other smart toilet user and the smart toilet user corresponding to the first central preference feature in the overall graph of smart toilet users, wherein the connectivity is related to the path length of the shortest traversal path of the smart toilet user in the overall graph of smart toilet users; Based on the corresponding transfer level, traverse each other smart toilet user in turn; For each other smart toilet user currently traversed, based on the transferred user preference features of the smart toilet user of the previous transfer level connected to the other smart toilet user, the current feature transfer is performed on the user preference features to be transferred corresponding to the other smart toilet user to form the transferred user preference features.

4. The method for pushing advertisements for smart toilets according to claim 3, characterized in that: The step of performing current feature transfer on the user preference features to be transferred corresponding to each other smart toilet user currently traversed based on the transferred user preference features of the smart toilet user of the previous transfer level connected to the other smart toilet user to form the transferred user preference features includes: For each other smart toilet user currently traversed, the association parameters between the transferred user preference features of each smart toilet user at the previous transfer level connected to the other smart toilet user and the user preference features to be transferred corresponding to the other smart toilet user are determined respectively, and based on the association parameters, the user preference features to be transferred corresponding to the other smart toilet user are adjusted, and the corresponding adjusted user preference features are output; For each other smart toilet user currently traversed, the mean of each adjusted user preference feature corresponding to the other smart toilet user is calculated to obtain the corresponding mean user preference feature, and the mean user preference feature and the user preference feature to be transmitted corresponding to the other smart toilet user are superimposed to obtain the transmitted user preference feature.

5. The method for pushing advertisements for smart toilets according to claim 1, characterized in that: The step of determining, for each of the time intervals, based on the first preference aggregation feature and the second preference aggregation feature corresponding to the time interval, at least one candidate advertisement matching the time interval from a plurality of candidate advertisements as the corresponding target preference advertisement includes: For each of the time intervals, feature aggregation is performed on the user preference features corresponding to each smart toilet user in the smart toilet user graph corresponding to the time interval to obtain a third preference aggregation feature corresponding to the time interval, and the first preference aggregation feature, the second preference aggregation feature and the third preference aggregation feature corresponding to the time interval are averaged to obtain a target preference aggregation feature corresponding to the time interval; For each candidate advertisement among the multiple candidate advertisements, feature mining is performed on the candidate advertisement to obtain advertisement content features corresponding to the candidate advertisement, and feature correlations between the advertisement content features and target preference aggregation features corresponding to each time interval are determined respectively to obtain push matching parameters between the candidate advertisement and each time interval; For each of the time intervals, based on the push matching parameters between the time interval and each of the candidate advertisements, at least one candidate advertisement matching the time interval is determined from among the multiple candidate advertisements as the corresponding target preference advertisement.

6. An advertising push system for smart toilets, characterized in that: include: A toilet user graph construction module is used to construct a smart toilet user graph corresponding to each time interval based on each smart toilet user who uses the smart toilet in the time interval and the user preference representation data corresponding to each smart toilet user, wherein in the smart toilet user graph, any two connected smart toilet users have a correlation relationship based on the corresponding user preference representation data, and the user preference representation data is used to reflect the behavioral tendencies of the smart toilet users; A toilet user graph connection module is used to connect the smart toilet users that match between every two smart toilet user graphs in a plurality of smart toilet user graphs corresponding to a plurality of time intervals to form a corresponding smart toilet user total graph, wherein the matched smart toilet users refer to the same smart toilet users; a target preference advertisement determination module, for each of the time intervals, based on the smart toilet user graph corresponding to the time interval and the smart toilet user total graph, to determine at least one candidate advertisement matching the time interval from a plurality of candidate advertisements as the corresponding target preference advertisement, wherein the advertising device in the smart toilet is used to play the corresponding target preference advertisement in the corresponding time interval; Wherein, for each of the time intervals, based on the smart toilet user graph corresponding to the time interval and the total smart toilet user graph, at least one candidate advertisement matching the time interval is determined from multiple candidate advertisements as the corresponding target preference advertisement, including: Performing feature mining on the user preference representation data corresponding to each smart toilet user in the overall map of smart toilet users to obtain the user preference features corresponding to each smart toilet user; The user preference feature corresponding to each smart toilet user in the total smart toilet user map is transferred to obtain a first user preference transfer feature corresponding to each smart toilet user, and, for each time interval, the user preference feature corresponding to each smart toilet user in the smart toilet user map corresponding to the time interval is transferred to obtain a second user preference transfer feature corresponding to each smart toilet user; For each of the time intervals, feature aggregation is performed on the first user preference transfer feature corresponding to each smart toilet user in the smart toilet user graph corresponding to the time interval to obtain the first preference aggregation feature corresponding to the time interval, and feature aggregation is performed on the second user preference transfer feature corresponding to each smart toilet user in the smart toilet user graph corresponding to the time interval to obtain the second preference aggregation feature corresponding to the time interval; For each of the time intervals, based on the first preference aggregation feature and the second preference aggregation feature corresponding to the time interval, at least one candidate advertisement matching the time interval is determined from a plurality of candidate advertisements as the corresponding target preference advertisement; The method of transferring the user preference features corresponding to each smart toilet user in the overall smart toilet user graph to obtain the first user preference transfer features corresponding to each smart toilet user, and transferring the user preference features corresponding to each smart toilet user in the smart toilet user graph corresponding to the time interval to obtain the second user preference transfer features corresponding to each smart toilet user, includes: Clustering the user preference features corresponding to each smart toilet user in the overall smart toilet user map to form at least one first central preference feature, and based on the at least one first central preference feature, performing feature transfer on the user preference features corresponding to each smart toilet user in the overall smart toilet user map to obtain a corresponding first user preference transfer feature; For each of the time intervals, clustering is performed on the user preference features corresponding to each smart toilet user in the smart toilet user graph corresponding to the time interval to form at least one second central user preference feature, and based on the at least one second central user preference feature, feature transfer is performed on the user preference features corresponding to each smart toilet user in the smart toilet user graph corresponding to the time interval to obtain a corresponding second user preference transfer feature; The method of clustering the user preference features corresponding to each smart toilet user in the overall smart toilet user map to form at least one first center preference feature, and based on the at least one first center preference feature, performing feature transfer on the user preference features corresponding to each smart toilet user in the overall smart toilet user map to obtain the corresponding first user preference transfer feature includes: Counting the smart toilet users in the overall map of smart toilet users to obtain the number of user statistics corresponding to the overall map of smart toilet users, and determining the number of cluster centers based on the number of user statistics, wherein the number of cluster centers is less than the number of user statistics; Based on the number of cluster centers, clustering the user preference features corresponding to each smart toilet user in the overall map of smart toilet users to form a corresponding number of first center preference features, and based on the number of user preference features in the cluster clusters corresponding to each of the first center preference features, determining the transmission priority corresponding to each of the first center preference features; Based on the corresponding delivery priority, traverse each of the first center preference features in turn; Taking the first central preference feature currently traversed as the starting point, the current feature transfer is performed on the user preference feature to be transferred corresponding to each smart toilet user in the total map of smart toilet users to form a transferred user preference feature, wherein the transferred user preference feature formed by the current feature transfer is used as the user preference feature to be transferred in the next feature transfer; The transferred user preference feature formed by performing feature transfer starting from the last traversed first central preference feature is used as the corresponding first user preference transfer feature.

7. An electronic device, characterized in that: include: Memory for storing computer programs; A processor connected to the memory is used to execute a computer program stored in the memory to implement the advertising push method for a smart toilet as described in any one of claims 1-5.

Images