Intelligent scene generation method, device, system, storage medium and electronic device

Abstract

This invention provides a method, apparatus, system, storage medium, and electronic device for generating intelligent scenes. The method includes: upon detecting a state change in a first device, acquiring a device state change message associated with the first device, the device state change message indicating the state change of the first device; generating a device list based on the device state change message, the device list including at least one device identifier corresponding to a device allowed to generate an intelligent scene; in response to an interactive operation performed on the device list, selecting a target device; and generating a target intelligent scene based on the target device, the first device, and the state change of the first device. This invention solves the technical problem in related technologies where the intelligent scene generation steps are cumbersome, leading to low efficiency in creating intelligent scenes.

Description

Technical Field

[0001] The embodiments of the present invention relate to the field of computers, and more specifically, to a method, apparatus, system, storage medium, and electronic device for generating intelligent scenes. Background Technology

[0002] With the rapid popularization of mobile devices and the high-speed development of cloud services, smart home applications are becoming increasingly widespread. Most smart home apps support automatic device control, one-click execution, and other smart scenarios. These apps typically present all of the user's devices directly to the user, allowing the user to select the conditions for execution and the actions to be performed by the devices. The app then sends the instructions to the platform, which in turn sends them to the devices for execution.

[0003] However, most users may not know where to start when encountering the series of functions of intelligent scenes. The main reason is that intelligent scenes require a lot of setup operations, and no guidance is provided to users, making the generation process of intelligent scenes cumbersome, which in turn leads to the technical problem of low generation efficiency of intelligent scenes.

[0004] There is currently no effective solution to the technical problem of low efficiency in generating intelligent scenes due to the cumbersome steps involved. Summary of the Invention

[0005] This invention provides a method, apparatus, system, storage medium, and electronic device for generating intelligent scenes, in order to at least solve the technical problem that the generation of intelligent scenes is cumbersome and inefficient.

[0006] According to an embodiment of the present invention, a method for generating a smart scene is provided, comprising: upon detecting a state change in a first device, acquiring a device state change message associated with the first device, wherein the device state change message is used to indicate the state change that has occurred in the first device; generating a device list based on the device state change message, wherein the device list includes at least one device identifier, the device identifier corresponding to a device that is allowed to generate a smart scene; in response to an interactive operation performed on the device list, selecting a target device, and generating a target smart scene based on the target device, the first device, and the state change that has occurred in the first device, wherein the target device, the first device, and the state change that has occurred in the first device are automatically associated in the target smart scene.

[0007] According to another embodiment of the present invention, an apparatus for generating a smart scene is provided, comprising: an acquisition module, which, upon detecting a state change in a first device, acquires a device state change message associated with the first device, wherein the device state change message is used to indicate a state change that has occurred in the first device; a generation module, which generates a device list based on the device state change message, wherein the device list includes at least one device identifier, the device identifier corresponding to a device that is allowed to generate a smart scene; and a processing module, which, in response to an interactive operation performed on the device list, selects a target device and generates a target smart scene based on the target device, the first device, and the state change that has occurred in the first device, wherein the target device, the first device, and the state change that has occurred in the first device are automatically associated in the target smart scene.

[0008] Optionally, the generation module includes: a first determining unit, configured to determine a first tail entity vector based on the device state change message, wherein the first device corresponds to a first head entity vector, the state change of the first device corresponds to a first relation vector, and the first head entity vector, the first relation vector, and the first tail entity vector together constitute a first triplet in a pre-determined intelligent scene triplet set; and a first generation unit, configured to generate the device list based on the first tail entity vector, wherein the similarity between the tail entity vectors corresponding to the devices in the device list and the first tail entity vector satisfies a preset condition.

[0009] Optionally, the first determining unit includes: a first extraction subunit, configured to extract the first head entity vector and the first relation vector according to the device state change message; a first processing subunit, configured to input the first head entity vector and the first relation vector into a pre-trained target neural network model to obtain the first tail entity vector, wherein the target neural network model is a model obtained by training an initial neural network model to be trained using a set of sample triples, the set of sample triples includes a set of positive sample triples and a set of negative sample triples, the set of positive sample triples is a set of labeled triples, and the set of negative sample triples is a set of triples obtained by replacing the head entity vector or the tail entity vector of the triples in the set of positive sample triples; the first determining subunit is configured to iteratively train the initial neural network model to obtain the first tail entity vector. During the initial neural network modeling process, when the input to the initial neural network model is a negative sample triplet from the negative sample triplet set, and the distance between the sum vector formed by the negative sample head entity vector and the negative sample relation vector and the negative sample tail entity vector is greater than a first distance threshold, and / or, when the input to the initial neural network model is a positive sample triplet from the positive sample triplet set, and the distance between the sum vector formed by the positive sample head entity vector and the positive sample relation vector and the positive sample tail entity vector is less than a second distance threshold, it is confirmed that the loss function of the initial neural network model satisfies a preset loss condition, and the initial neural network model is determined as the target neural network model; when the loss function does not satisfy the preset loss condition, the parameters of the initial neural network model are adjusted until the loss function satisfies the preset loss condition.

[0010] Optionally, the generation module includes: a first acquisition unit, configured to acquire historical data of device state changes, wherein the historical data of device state changes includes a second device that has undergone a state change, first time information of the second device undergoing a state change, a third device that has undergone a state change within a preset time interval after the first time information, and second time information of the third device undergoing a state change; and a first construction unit, configured to construct a second triplet in the intelligent scene triplet set based on the historical data of device state changes, wherein the second triplet includes a second head entity vector corresponding to the second device, a second relation vector corresponding to the state change of the second device, and a second tail entity vector corresponding to the third device.

[0011] Optionally, the generation module further includes: a second construction unit, configured to construct a third triplet in the intelligent scene triplet set based on the device state change history data when the device state change history data does not include the third device and the second time information, wherein the third triplet includes an empty third head entity vector, a third relation vector corresponding to the second time information, and a third tail entity vector corresponding to the second device; or, a third construction unit, configured to construct a fourth triplet in the intelligent scene triplet set based on the device state change history data when the first time information and the second time information included in the device state change history data indicate that the second device and the third device need to keep their state changes synchronized, wherein the fourth triplet includes a second head entity vector, a fourth relation vector indicating that the second device and the third device need to keep their state changes synchronized, and a second tail entity vector.

[0012] Optionally, the generation module further includes: a second acquisition unit, configured to acquire a set of tail entity vectors whose distance values ​​with the first tail entity vector satisfy a preset distance threshold, wherein the sorting position of the tail entity vectors in the set of tail entity vectors is negatively correlated with the distance value; and a second generation unit, configured to generate the device list based on the set of tail entity vectors, wherein the devices in the device list are arranged according to the sorting position of the tail entity vectors.

[0013] Optionally, the generation module further includes: a first display unit, configured to display a target prompt message on a target client, wherein the target prompt message is used to prompt that the first device has undergone a state change and that the generation of a smart scene associated with the first device is currently permitted; a second display unit, configured to display a list of state changes of the first device in response to a confirmation operation performed on the target prompt message, wherein the list of state changes includes state changes permitted for the first device; and a third generation unit, configured to generate the device list based on the selected target state change in response to a selection operation performed on the list of state changes.

[0014] According to another embodiment of the present invention, a smart scene generation system is provided, characterized in that it includes:

[0015] A detection device is used to acquire a device state change message associated with the first device when a state change is detected in the first device, wherein the device state change message is used to indicate the state change that has occurred in the first device;

[0016] A server is configured to generate a device list based on device status change messages, wherein the device list includes at least one device identifier, the device identifier corresponding to a device that is allowed to generate a smart scene;

[0017] An application is configured to select a target device in response to an interactive operation performed on the device list, and generate a target smart scene based on the target device, the first device, and the state changes that occur to the first device, wherein the target device, the first device, and the state changes that occur to the first device are automatically associated in the target smart scene.

[0018] According to yet another embodiment of the present invention, a computer-readable storage medium is also provided, wherein a computer program is stored therein, wherein the computer program is configured to perform the steps in any of the above method embodiments when executed.

[0019] According to yet another embodiment of the present invention, an electronic device is also provided, including a memory and a processor, wherein the memory stores a computer program and the processor is configured to run the computer program to perform the steps in any of the above method embodiments.

[0020] This invention addresses the technical problem of cumbersome smart scene generation steps and low efficiency in related technologies. By acquiring a device state change message associated with the first device, indicating the state change, upon detecting a state change in the first device, and then generating a device list including at least one device identifier based on the device state change message (the device identifiers correspond to devices allowed to generate smart scenes), and in response to an interactive operation on the device list, selecting a target device, and generating a target smart scene based on the target device, the first device, and the state change of the first device, where the target device, the first device, and the state change of the first device are automatically associated within the target smart scene, this invention achieves the technical effect of improving the efficiency of smart scene generation, providing users with appropriate operational guidance, and simplifying the smart scene generation method. Attached Figure Description

[0021] Figure 1 This is a hardware structure block diagram of a mobile terminal for a method of generating intelligent scenes according to an embodiment of the present invention.

[0022] Figure 2 This is a flowchart of a method for generating an intelligent scene according to an embodiment of the present invention;

[0023] Figure 3 This is a schematic diagram illustrating a specific example of a method for generating an intelligent scene according to an embodiment of the present invention;

[0024] Figure 4 This is a flowchart illustrating the model generation process of a method for generating an intelligent scene according to an embodiment of the present invention.

[0025] Figure 5 This is a flowchart of the model training process for a method of generating an intelligent scene according to an embodiment of the present invention.

[0026] Figure 6 This is a model training flowchart of another intelligent scene generation method according to an embodiment of the present invention;

[0027] Figure 7 This is a flowchart of another method for generating intelligent scenes according to an embodiment of the present invention;

[0028] Figure 8 This is a flowchart of another method for generating intelligent scenes according to an embodiment of the present invention;

[0029] Figure 9 This is a structural block diagram of an intelligent scene generation device according to an embodiment of the present invention. Detailed Implementation

[0030] The embodiments of the present invention will be described in detail below with reference to the accompanying drawings and examples.

[0031] It should be noted that the terms "first," "second," etc., in the specification, claims, and drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence.

[0032] The methods and embodiments provided in this application can be executed on a mobile terminal, computer terminal, or similar computing device. Taking running on a mobile terminal as an example, Figure 1 This is a hardware structure block diagram of a mobile terminal for a method of generating intelligent scenes according to an embodiment of the present invention. For example... Figure 1 As shown, a mobile terminal may include one or more ( Figure 1 Only one is shown in the diagram. A processor 102 (which may include, but is not limited to, a microprocessor MCU or a programmable logic device FPGA, etc.) and a memory 104 for storing data are also shown. The mobile terminal may further include a transmission device 106 for communication functions and an input / output device 108. Those skilled in the art will understand that... Figure 1 The structure shown is for illustrative purposes only and does not limit the structure of the mobile terminal described above. For example, the mobile terminal may also include components that are more... Figure 1 The more or fewer components shown, or having the same Figure 1 The different configurations shown.

[0033] The memory 104 can be used to store computer programs, such as application software programs and modules, like the computer program corresponding to the intelligent scene generation method in this embodiment of the invention. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, thereby implementing the above-described method. The memory 104 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include memory remotely located relative to the processor 102, and these remote memories can be connected to the mobile terminal via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0034] The transmission device 106 is used to receive or send data via a network. Specific examples of the network described above may include a wireless network provided by the mobile terminal's communication provider. In one example, the transmission device 106 includes a Network Interface Controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the transmission device 106 may be a Radio Frequency (RF) module, used for wireless communication with the Internet.

[0035] This embodiment provides a method for generating intelligent scenes. Figure 2 This is a flowchart of a method for generating intelligent scenes according to an embodiment of the present invention, such as... Figure 2 As shown, the process includes the following steps:

[0036] S202, when a state change of the first device is detected, a device state change message associated with the first device is obtained, wherein the device state change message is used to indicate the state change of the first device;

[0037] Optionally, in this embodiment, the first device may include, but is not limited to, devices that allow the generation of intelligent scenes and the acquisition of state change messages, such as: lights, curtains, door locks, electrical appliances (televisions, air conditioners, humidifiers, water dispensers, water heaters, refrigerators, range hoods, washing machines, dryers, microwave ovens, ovens), robot vacuum cleaners, floor heating, smoke detectors, etc.

[0038] Optionally, in this embodiment, the aforementioned state changes may include, but are not limited to, state adjustments or changes that allow the device generating the smart scene to undergo, such as unlocking (closing) the door, turning the lights on (off), adjusting the air conditioner temperature, turning the robot vacuum cleaner on (off), and opening (closing) the curtains.

[0039] Optionally, in this embodiment, the device state change message may include, but is not limited to, the message generated after the state of the device that allows the generation of smart scenes changes. The device state change message is used to indicate the state change of the first device. For example, when the user opens or closes the door lock, the platform sends the following message to the client: The door lock is open (closed).

[0040] S204, Generate a device list based on the device status change message, wherein the device list includes at least one device identifier, and the device identifier corresponds to the device that is allowed to generate a smart scene;

[0041] Optionally, in this embodiment, the device list may include, but is not limited to, a set of devices that allow the generation of smart scenes based on device status change messages, and the set of devices may include some or all of the devices in the smart scene.

[0042] Optionally, in this embodiment, the device identifier may include, but is not limited to, devices that allow the generation of smart scenes. Users can select desired devices based on different device identifiers, such as device names (refrigerators, air conditioners, washing machines, door locks, etc.), graphics, animations, photos, and sounds representing the device, and may also include, but is not limited to, one or more combinations of the above.

[0043] S206, in response to the interactive operation performed on the device list, the target device is selected, and a target intelligent scene is generated based on the target device, the first device, and the state changes that occur in the first device, wherein the target device, the first device, and the state changes that occur in the first device are automatically associated in the target intelligent scene.

[0044] Optionally, in this embodiment, the target device may include, but is not limited to, the device selected from the device list. For example, after the client receives a device list pushed by the platform (the list includes devices such as robot vacuums, air conditioners, and water heaters), the user selects an air conditioner from the list, and then the air conditioner is the target device.

[0045] Optionally, in this embodiment, the aforementioned target intelligent scene may include, but is not limited to, an intelligent scene generated from the target device selected from the device list generated based on the state changes of the first device, the first device, and the state changes that have occurred in the first device.

[0046] For example, Figure 3 This is a schematic diagram of a method for generating an intelligent scene according to an embodiment of the present invention, such as... Figure 3As shown: When a user unlocks the door, the action 302 is recognized by the system as a change in the door lock's status. Message 306 pops up on the client interface 304, informing the user: "The door lock has been opened. Most users will operate devices such as air conditioners, lights, curtains, water heaters, and floor heating after unlocking the door. It is recommended to create a smart scene." In response to the user's interaction 310, the system performs an interaction on the device list 308 pushed by the client, selecting the air conditioner in the device list 308, turning on the air conditioner, and adjusting its operating properties. At this point, the door lock, unlocking, and air conditioner can generate a target smart scene.

[0047] The above is merely an example, and this application does not impose any specific limitations.

[0048] This application's embodiments employ a method that, upon detecting a state change in a first device, obtains a device state change message associated with the first device, generates a device list based on the device state change message, and the device list includes at least one device identifier, which corresponds to a device allowed to generate a smart scene. In response to an interactive operation performed on the device list, a target device is selected, and a target smart scene is generated based on the target device, the first device, and the state change of the first device. The automatic association of the target device, the first device, and the state change of the first device within the target smart scene solves the technical problem of cumbersome smart scene generation steps and low efficiency in related technologies. This achieves the technical effects of improving smart scene generation efficiency, providing users with suitable operational guidance, and simplifying the smart scene generation method.

[0049] In an exemplary embodiment, generating a device list based on a device state change message includes: determining a first tail entity vector based on the device state change message, wherein a first device corresponds to a first head entity vector, the state change of the first device corresponds to a first relation vector, and the first head entity vector, the first relation vector, and the first tail entity vector together constitute a first triplet in a pre-determined intelligent scene triplet set; generating a device list based on the first tail entity vector, wherein the similarity between the tail entity vectors corresponding to the devices in the device list and the first tail entity vector satisfies a preset condition.

[0050] Optionally, in this embodiment, the first head entity vector may include, but is not limited to, the vector composed of the first device whose state has changed, and the first relation vector may include, but is not limited to, the vector composed of the first device whose state has changed.

[0051] Specifically, each action can be broken down into head entities and relations. This can include, but is not limited to, using word vectors pre-trained based on word2vec to represent head entities and relations, or using one-hot encoding to represent head entities and relations, thereby determining the aforementioned first head entity vector and first relation vector.

[0052] It should be noted that word2vec and one-hot encoding mentioned above are specific models in word encoding models. In word2vec, the word vectors are pre-trained, and the representation vectors corresponding to the specific words and sentences can be obtained by inputting them into the encoding model.

[0053] For example, a user's fingerprint unlocking action can be broken down into the head entity "lock" and the relation "fingerprint unlocking". Then, the head entity "lock" and the relation "fingerprint unlocking" can be represented by vectors based on word vectors pre-trained by word2vec, so as to obtain the first head entity vector corresponding to the head entity "lock" and the first relation vector corresponding to the fingerprint unlocking. Alternatively, the head entity and the relation can be represented by vectors through one-hot encoding.

[0054] The above is merely an example, and this application does not impose any specific limitations.

[0055] Optionally, in this embodiment, the first triplet is composed of a first head entity vector, a first relation vector, and a first tail entity vector, and the intelligent scene triplet set is a set composed of multiple triplets.

[0056] Optionally, in this embodiment, the method for determining the first tail entity vector based on the device state change may include, but is not limited to, training a neural network model, inputting the first head entity vector and the first relation vector into the neural network model to obtain the first tail entity vector. The neural network model is pre-trained and is used to predict the first tail entity vector corresponding to the first head entity vector and the first relation vector.

[0057] Specifically, as a type of neural network model, the transE model can be trained by inputting numerous triplet data consisting of head entities, relations, and tail entities into the model. Then, the first head entity vector and the first relation vector corresponding to the head entity and relation are input into the trained transE model, and the aforementioned first tail entity vector is output.

[0058] It's important to note that transE's design philosophy is similar to vector summation in linear algebra. The basic idea behind transE training is to represent entities and relationships as m-dimensional vectors, such as a head entity vector h, a tail entity vector t, and a relationship vector l. Then, for head entity vectors h, tail entity vector t, and relationship vector l that actually have a relationship, a relationship of h + l ≈ t is allowed. That is, the closer h + l is to t, the smaller the deviation, and the greater the likelihood that the user will choose a device from the device list corresponding to the tail entity vector; conversely, the less likely the user is to choose a device from the device list corresponding to the tail entity vector.

[0059] For example, Figure 4 This is a flowchart illustrating an example of a method for generating an intelligent scene according to the present invention, such as... Figure 4 As shown, the flowchart includes the following steps:

[0060] S402, Begin;

[0061] S404, decompose the user's execution action into device entity (i.e., the head entity corresponding to the above) and associated action (i.e., the relationship corresponding to the above) for subsequent model training;

[0062] S406, convert the device entity into a first head entity vector representation, and convert the associated action into a first relation phasor representation;

[0063] S408, input the transformed first head entity vector and first relation vector into the transE model;

[0064] S410, the transE model makes predictions based on the first head entity vector and the first relation vector;

[0065] S412, outputs the first tail entity vector;

[0066] S414, End.

[0067] Optionally, in this embodiment, the similarity may include, but is not limited to, cosine similarity. The degree of similarity between two vectors is determined by calculating the cosine distance between them. The closer the cosine distance between two vectors is to 1, the more similar the two vectors are. In this case, the user is more likely to generate the above-mentioned intelligent scene based on the device corresponding to the vector. The above-mentioned preset condition can be understood as the cosine distance meeting a preset distance threshold.

[0068] In an exemplary embodiment, determining the first tail entity vector based on the device state change message includes:

[0069] Extract the first entity vector and the first relation vector based on the device status change message;

[0070] The first head entity vector and the first relation vector are input into the pre-trained target neural network model to obtain the first tail entity vector. If the parameters of the target neural network do not meet the preset loss condition, the parameters of the initial neural network model are adjusted until the loss model is a model obtained by training the initial neural network model to be trained using the sample triplet set. The sample triplet set includes a positive sample triplet set and a negative sample triplet set. The positive sample triplet set is a set of labeled triplets, and the negative sample triplet set is a set of triplets obtained by replacing the head entity vector or the tail entity vector of the triplet in the positive sample triplet set.

[0071] During the iterative training of the initial neural network model, if the input to the initial neural network model is a negative triplet from the set of negative triplets, and the distance between the sum vector of the negative sample head entity vector and the negative sample relation vector and the negative sample tail entity vector is greater than a first distance threshold, and / or if the input to the initial neural network model is a positive triplet from the set of positive triplets, and the distance between the sum vector of the positive sample head entity vector and the positive sample relation vector and the positive sample tail entity vector is less than a second distance threshold, then the loss function of the initial neural network model is confirmed to meet the preset loss condition, and the initial neural network model is determined as the target neural network model; if the loss function does not meet the preset loss condition, the parameters of the initial neural network model are adjusted until the loss function meets the preset loss condition.

[0072] Optionally, in this embodiment, a triple is composed of a head entity, a relation, and a tail entity, and can be represented as <head entity, relation, tail entity>. The sample triple set includes a positive sample triple set and a negative sample triple set. The positive sample triple set can include, but is not limited to, a set of triples that are actually related. For example, if a user turns on the light every time they unlock the door, then the triple <door lock, unlock, light> is indeed related. This triple is the positive sample triple set mentioned above. The set of triples composed of similar triples is the positive sample triple set mentioned above. Alternatively, the positive sample vector set can also include, but is not limited to, a set of triples that are pre-annotated based on experience, or a combination of the above schemes.

[0073] Optionally, in this embodiment, the preset loss condition can be a condition pre-set according to requirements or a condition pre-set according to previous neural network training experience. In addition, when the loss function does not meet the preset loss condition, the parameters of the initial neural network model can be manually adjusted or the model can automatically adjust the parameters of the neural network model.

[0074] Optionally, in this embodiment, the distance threshold can be a range of preset deviation values ​​from the sum of the head entity vector and the relation vector to the tail entity vector. The distance threshold can also be a value determined based on historical experience. For example, when the preset first distance threshold is 2 and the second distance threshold is 3, the distance between the sum vector formed by the negative sample head entity vector and the negative sample relation vector and the negative sample tail entity vector is greater than 2, and / or when the input to the initial neural network model is a positive sample triplet in the set of positive sample triplets, the distance between the sum vector formed by the positive sample head entity vector and the positive sample relation vector and the positive sample tail entity vector is less than the second distance threshold 3, it is confirmed that the loss function of the initial neural network model satisfies the preset loss condition.

[0075] Specifically, h represents the first head entity vector, l represents the first relation vector, and t represents the first tail entity vector. The distance function d is defined as the deviation from h+l to t, i.e.

[0076]

[0077] The purpose of this model is to shrink the correct triplet d and amplify the incorrect triplet d, hence the loss function L, where h′ represents the replaced head entity vector, t′ represents the replaced tail entity vector, and S represents the set of triplets:

[0078]

[0079] By calculating d, we determine whether d is greater than the first distance threshold or less than the second distance threshold to determine whether the parameters of the initial neural network model meet the preset loss condition. If the result is yes, the condition is met, and the initial neural network model is determined to be the target neural network model. If the result is no, the preset loss condition is not met, the parameters of the initial neural network model are adjusted, and iterative training continues until the result is yes.

[0080] The above is merely a specific example, and this application does not impose any limitations.

[0081] Optionally, in this embodiment, the neural network model is implemented by adjusting the initial loss model parameters, defining a distance function to represent the deviation between the first head entity vector and the first tail entity vector, until the loss model is a model obtained by training the initial neural network model to be trained using the sample triplet set.

[0082] For example, Figure 5 This is a flowchart of neural network model training, such as... Figure 5 As shown, the steps for training a neural network model may include, but are not limited to:

[0083] S502, Input the set of sample triples into the initial neural network model;

[0084] S504, calculate the distance function d, where d represents the deviation from the sum of the first head entity vector and the first relation vector to the first tail entity vector;

[0085] S506, Determine whether d meets the distance threshold condition;

[0086] S506-1, if the result is yes, execute S508, then the initial neural network model is the target neural network model;

[0087] S506-2, if the result is negative, execute S510 to adjust the initial neural network model parameters;

[0088] S508 outputs the target neural network model;

[0089] S510, adjust the initial neural network model parameters.

[0090] In an exemplary embodiment, generating a device list based on device status change messages includes: obtaining historical device status change data, wherein the historical device status change data includes a second device that has undergone a status change, first time information of the second device undergoing a status change, a third device that undergoes a status change within a preset time interval after the first time information, and second time information of the third device undergoing a status change; constructing a second triplet in the intelligent scene triplet set based on the historical device status change data, wherein the second triplet includes a second head entity vector corresponding to the second device, a second relation vector corresponding to the status change of the second device, and a second tail entity vector corresponding to the third device.

[0091] Optionally, in this embodiment, the aforementioned historical data may be obtained from the data of other users' execution actions in the database, or it may be the historical execution action data of users stored in the database, or it may be historical data preset by the system, or it may be a combination of the above, including but not limited to. The execution action records of the current user or other users can be obtained based on the historical data to collect the aforementioned user execution action data. The data content recorded in the historical data may include, but is not limited to, the identifier of the second device whose state changed, the first time the second device changed, the third device whose state changed within a preset time interval after the first time, and the second time information of the third device's state change.

[0092] Optionally, in this embodiment, the above preset time interval can be 1 second or 1 minute, and it can be set manually or preset by the system according to past experience. For example, if the preset time interval is 2 minutes, and the user opens the bedroom curtains at 8:00 every morning and closes the door lock at 8:05 in the morning, and the time interval in between is 5 minutes, then this data cannot be used as historical data. If the user opens the bedroom curtains at 8:00 every morning and closes the door lock at 8:01 in the morning, then this data can be used as historical data, and this historical data will be used to construct the second triple in the intelligent scenario triple set to ultimately construct the above intelligent scenario triple set.

[0093] The above is only a specific example, and this application makes no limitations.

[0094] In an exemplary embodiment, the above method further includes:

[0095] In the case where the device status change historical data does not include the third device and the second time information, constructing the third triple in the intelligent scenario triple set according to the device status change historical data, where the third triple includes an empty third head entity vector, a third relation vector corresponding to the second time information, and a third tail entity vector corresponding to the second device; or,

[0096] In the case where the first time information and the second time information included in the device status change historical data indicate that the second device and the third device need to maintain a synchronous status change, constructing the fourth triple in the intelligent scenario triple set according to the device status change historical data, where the fourth triple includes a second head entity vector, a fourth relation vector indicating that the second device and the third device need to maintain a synchronous status change, and a second tail entity vector.

[0097] Optionally, in this embodiment, the above third triple can be expressed as <null, timing, device>. For example, if the execution action is that the user opens the curtains at 8:00 every morning and the user does not perform other execution actions within the preset interval before opening the curtains, then the triple corresponding to the execution action of the user opening the curtains at 8:00 every morning should be <null, 8:00 am, curtains>, so as to ultimately construct the above intelligent scenario triple set.

[0098] Optionally, in this embodiment, the fourth triple can be expressed as <device A, synchronous, device Z>. For example, if the execution action is that the user often turns on the humidifier while turning on the air conditioner, then the execution action of the user turning on the humidifier while turning on the air conditioner can be expressed as the corresponding triple <air conditioner, synchronous, humidifier>, and this triple is the fourth triple in the triple set.

[0099] In one exemplary embodiment, generating a device list based on the first tail entity vector includes:

[0100] Obtain a set of tail entity vectors whose distance values ​​from the first tail entity vector satisfy a preset distance threshold, wherein the sorting position of the tail entity vectors in the set of tail entity vectors is negatively correlated with the distance value;

[0101] A device list is generated based on the set of tail entity vectors, where the devices in the device list are arranged according to the sorting position of the tail entity vectors.

[0102] Optionally, in this embodiment, the sorting position of the tail entity vectors in the tail entity vector set is negatively correlated with the corresponding distance value. This can be achieved by inputting the head entity vector and relation vector into the neural network training model to obtain the tail entity vectors, and then sorting the tail entity vectors in the tail entity vector set with the first tail entity vector obtained by inputting it into the neural network model.

[0103] It should be noted that, for the distance function d defined above, the greater the distance from h+l to t, the later the vector of t is ordered. For example, if the distance from the light and the light being turned on to the curtains being turned on is 1, and the distance from the light and the light being turned on to the air conditioner being turned on is 2, then the order of the tail vectors corresponding to the curtains and the air conditioner in the tail vector set should be: curtains, air conditioner.

[0104] In an exemplary embodiment, generating a device list based on a device state change message includes: displaying a target prompt message on a target client, wherein the target prompt message is used to prompt that a first device has undergone a state change and that the generation of a smart scene associated with the first device is currently permitted; displaying a state change list of the first device in response to a confirmation operation performed on the target prompt message, wherein the state change list includes state changes that the first device is permitted to perform; and generating a device list based on the selected target state change in response to a selection operation performed on the state change list.

[0105] Optionally, in this embodiment, as Figure 3 As shown, the user unlocks the door. This action 302 is recognized by the system, and the door lock status changes. Message 306 pops up on the client interface 304. Message 306 informs the user that the door lock has been opened. Most users will operate devices such as air conditioners, lights, curtains, water heaters, and floor heating after unlocking the door, and it is recommended to create a smart scene. In response to the device list 308 pushed by the client, the user 310 performs an interactive action, clicks on message 306, selects the air conditioner in the device list 308, turns on the air conditioner, and adjusts the air conditioner's operating properties. Thus, the door lock, unlocking, and air conditioner generate a target smart scene.

[0106] Obviously, the embodiments described above are only some embodiments of the present invention, and not all embodiments.

[0107] The present application will be described in detail below with reference to specific embodiments:

[0108] This application primarily addresses the technical problem of cumbersome smart scene generation steps, leading to low efficiency in smart scene creation. Based on the user's current action, the transE method predicts the devices that might be operated next and pushes these predictions to the mobile app, prompting the user to navigate to a designated smart scene settings page (including one-click execution, automatic control, group control, etc.). Users can select from recommended devices or devices not found in the list, thus creating a behavior-guided approach, simplifying tedious setup, and improving the user experience.

[0109] like Figure 6 As shown, the steps for generating intelligent scenes may include, but are not limited to, the following: triplet collection, transE training, prediction and push, and app execution.

[0110] 1. Construct a triplet library for intelligent scene entity relationships:

[0111] During the data collection phase, the client reports data points (including but not limited to the aforementioned device status change messages) when the device status changes, and performs statistical analysis. The big data platform integrates the data and constructs an entity relationship triplet library. The statistical information required for device status changes includes: device model, device status change, status change time, and the device model and status change time at the next moment.

[0112] During the data cleaning phase, records with long time intervals between two device status changes are removed. For records where both fields are empty at the next time step, the k most frequent time points are selected, and non-conditional device status changes are labeled with triples and associated as synchronization. The filtered records are then integrated into a triple data structure, as shown in the table below:

[0113]

[0114]

[0115] 2. TransE training:

[0116] The design of transE is similar to the summation of vectors in linear algebra. The basic idea is to represent both entities and relations as m-dimensional vectors. A triple can be represented as a head entity vector h, a tail entity vector t, and a relation vector l. Therefore, for a positive triple (a triple that actually has the association), the relation h + l ≈ t holds. The initial vectors can use word vectors pre-trained by word2vec as input and output; if the vocabulary is small, one-hot encoding can also be used.

[0117] The transE dataset is constructed by replacing only the head entity or only the tail entity in the existing triplet dataset (corresponding to the aforementioned negative sample triplets) and combining them with the existing triplet dataset. This can be represented as:

[0118] S′ (h,l,t) ={(h′,l,t)|h′∈E}∪{(h,l,t′)|t′∈E}

[0119] Where S′ represents the dataset of transE, h′ refers to the replaced head entity, E refers to the original triple dataset, h refers to the head entity, l refers to the relation, t refers to the head entity, and t′ refers to the replaced tail entity.

[0120] Furthermore, a distance function d is defined to represent the deviation from h+l to t, i.e.

[0121]

[0122] The purpose of this model is to reduce the size of positive triples (d) and increase the size of negative triples (d). Therefore, a loss function L is defined, where γ is the initial neural network model parameter, h refers to the head entity, l refers to the relation, t refers to the head entity, h′ refers to the replaced head entity, and t′ refers to the replaced tail entity.

[0123]

[0124] The model uses stochastic gradient descent (SGD) to update the parameter γ, ensuring that the loss function value is within a preset range, thus completing the training of the transE model.

[0125] 3. Prediction and Push Notifications:

[0126] like Figure 7As shown, any action can be broken down into an entity and an association. The entity is converted into a corresponding head entity vector representation, and the associated action is converted into a tail entity vector representation. The head entity vector and tail entity vector converted from the entity and associated action are input into the transE training model to obtain the vector representation of the tail entity. The tail entity vector is converted into a corresponding entity device list, and the entity ranking is obtained by calculating similarity comparison. When the user performs an action, the head entity, association, and the obtained entity ranking (i.e., the list of devices under the current user account that exist in the tail entity list) corresponding to the action are pushed to the client.

[0127] For example, fingerprint door opening can be broken down into door lock and fingerprint unlock. The word vectors for door lock and fingerprint unlock are input into the model to obtain the vector representation of the tail entity. Possible entity rankings are obtained through cosine similarity comparison, which might be [light bulb, water heater, robot vacuum cleaner...]. When the user performs the door opening action, the platform sends a push message to the client. The push message contains the head entity, its associations, and a list of devices (light bulb, water heater, robot vacuum cleaner...) that exist in the tail entity list under the current user account.

[0128] 4. App execution:

[0129] like Figure 8 As shown, the client locally stores an enumeration of association types. When the client receives a push notification from the platform, it parses the association and matches it with the corresponding enumeration type. The enumeration types are divided into device status change association, timed association, and synchronous execution association. The client jumps to different function pages based on different associations. The client creates a prompt message to remind the user (e.g., most users will have other associated actions after performing a certain action; it is recommended to create a smart scene). When the user clicks the push notification, different operations will be performed based on the association type. In addition, the client parses the head entity and association in the push notification, automatically fills in the conditions, and displays the queried tail entity list in order to recommend the user to select subsequent operations.

[0130] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods according to the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0131] This embodiment also provides an intelligent scene generation device, which is used to implement the above embodiments and preferred embodiments; details already described will not be repeated. As used below, the term "module" can be a combination of software and / or hardware that implements a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0132] Figure 9 This is a structural block diagram of an intelligent scene generation device according to an embodiment of the present invention, such as... Figure 9 As shown, the device includes:

[0133] The acquisition module 902 is used to acquire device status change messages associated with the first device, wherein the device status change messages are used to indicate the status changes that have occurred in the first device;

[0134] The generation module 904 is used to generate a device list based on the device status change message, wherein the device list includes at least one device identifier, and the device identifier corresponds to the device that is allowed to generate a smart scene;

[0135] The processing module 906 is used to respond to the interactive operation performed on the device list, select the target device, and generate a target intelligent scene based on the target device, the first device, and the state changes that occur in the first device, wherein the target device, the first device, and the state changes that occur in the first device are automatically associated in the target intelligent scene.

[0136] The generation module 904 includes:

[0137] The first determining unit is configured to determine the first tail entity vector based on the device state change message, wherein the first device corresponds to the first head entity vector, the state change of the first device corresponds to the first relation vector, and the first head entity vector, the first relation vector, and the first tail entity vector together constitute the first triplet in the pre-determined intelligent scene triplet set.

[0138] The first generation unit is configured to generate the device list based on the first tail entity vector, wherein the similarity between the tail entity vectors corresponding to the devices in the device list and the first tail entity vector satisfies a preset condition.

[0139] The first defined unit includes:

[0140] The first extraction subunit is used to extract the first header entity vector and the first relationship vector based on the device status change message.

[0141] The first processing subunit is used to input the first head entity vector and the first relation vector into a pre-trained target neural network model to obtain the first tail entity vector. The target neural network model is a model obtained by training an initial neural network model to be trained using a set of sample triples. The set of sample triples includes a set of positive sample triples and a set of negative sample triples. The set of positive sample triples is a set of labeled triples. The set of negative sample triples is a set of triples obtained by replacing the head entity vector or the tail entity vector of the triples in the set of positive sample triples.

[0142] The first determining subunit is configured to, during the iterative training of the initial neural network model, determine the initial neural network model as the target neural network model when, during the iterative training process, if the input to the initial neural network model is a negative sample triplet from the negative sample triplet set, and the distance between the sum vector formed by the negative sample head entity vector and the negative sample relation vector and the negative sample tail entity vector is greater than a first distance threshold, and / or when the input to the initial neural network model is a positive sample triplet from the positive sample triplet set, and the distance between the sum vector formed by the positive sample head entity vector and the positive sample relation vector and the positive sample tail entity vector is less than a second distance threshold, and / or confirm that the loss function of the initial neural network model satisfies a preset loss condition, and determine the initial neural network model as the target neural network model; when the loss function does not satisfy the preset loss condition, adjust the parameters of the initial neural network model until the loss function satisfies the preset loss condition.

[0143] Module 904 also includes:

[0144] The first acquisition unit is used to acquire historical data of device status changes, wherein the historical data of device status changes includes a second device that has undergone a status change, a first time information of the second device undergoing a status change, a third device that has undergone a status change within a preset time interval after the first time information, and a second time information of the third device undergoing a status change.

[0145] The first construction unit is used to construct a second triplet in the set of intelligent scene triplets based on the historical data of device state changes. The second triplet includes a second head entity vector corresponding to the second device, a second relation vector corresponding to the state change of the second device, and a second tail entity vector corresponding to the third device.

[0146] Generation module 904 also includes:

[0147] The second construction unit is configured to construct a third triplet in the intelligent scene triplet set based on the device state change history data when the device state change history data does not include the third device and the second time information. The third triplet includes an empty third head entity vector, a third relation vector corresponding to the second time information, and a third tail entity vector corresponding to the second device; or...

[0148] The third construction unit is used to construct a fourth triplet in the intelligent scene triplet set based on the device state change history data, when the first time information and the second time information included in the device state change history data indicate that the second device and the third device need to keep synchronized state changes. The fourth triplet includes a second head entity vector, a fourth relation vector indicating that the second device and the third device need to keep synchronized state changes, and a second tail entity vector.

[0149] Generation module 904 also includes:

[0150] The second acquisition unit is used to acquire a set of tail entity vectors whose distance values ​​with the first tail entity vector satisfy a preset distance threshold, wherein the sorting position of the tail entity vectors in the set of tail entity vectors is negatively correlated with the distance value;

[0151] The second generation unit is used to generate the device list based on the set of tail entity vectors, wherein the devices in the device list are arranged according to the sorting position of the tail entity vectors.

[0152] Generation module 904 also includes:

[0153] The first display unit is used to display a target prompt message on the target client, wherein the target prompt message is used to prompt that the first device has undergone a state change and that the generation of a smart scene associated with the first device is currently allowed;

[0154] The second display unit displays a list of state changes of the first device in response to a confirmation operation performed on the target prompt message, wherein the list of state changes includes state changes that the first device is allowed to perform.

[0155] The third generation unit generates the device list based on the selected target state change in response to a selection operation performed on the state change list.

[0156] According to another embodiment of the present invention, a smart scene generation system is also provided, comprising:

[0157] A detection device is used to acquire a device state change message associated with the first device when a state change is detected in the first device, wherein the device state change message is used to indicate the state change that has occurred in the first device;

[0158] A server is configured to generate a device list based on device status change messages, wherein the device list includes at least one device identifier, the device identifier corresponding to a device that is allowed to generate a smart scene;

[0159] An application is configured to select a target device in response to an interactive operation performed on the device list, and generate a target smart scene based on the target device, the first device, and the state changes that occur to the first device, wherein the target device, the first device, and the state changes that occur to the first device are automatically associated in the target smart scene.

[0160] According to yet another embodiment of the present invention, a computer-readable storage medium is also provided, wherein a computer program is stored therein, wherein the computer program is configured to perform the steps in any of the above method embodiments when executed.

[0161] It should be noted that the above modules can be implemented by software or hardware. For the latter, they can be implemented in the following ways, but are not limited to: all the above modules are located in the same processor; or, the above modules are located in different processors in any combination.

[0162] Embodiments of the present invention also provide a computer-readable storage medium storing a computer program, wherein the computer program is configured to perform the steps in any of the above method embodiments when executed.

[0163] In one exemplary embodiment, the aforementioned computer-readable storage medium may include, but is not limited to, various media capable of storing computer programs, such as a USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard disk, magnetic disk, or optical disk.

[0164] Embodiments of the present invention also provide an electronic device including a memory and a processor, the memory storing a computer program and the processor being configured to run the computer program to perform the steps in any of the above method embodiments.

[0165] In one exemplary embodiment, the electronic device may further include a transmission device and an input / output device, wherein the transmission device is connected to the processor and the input / output device is connected to the processor.

[0166] Specific examples in this embodiment can be found in the examples described in the above embodiments and exemplary implementations, and will not be repeated here.

[0167] It is obvious to those skilled in the art that the modules or steps of the present invention described above can be implemented using general-purpose computing devices. They can be centralized on a single computing device or distributed across a network of multiple computing devices. They can be implemented using computer-executable program code, and thus can be stored in a storage device for execution by a computing device. In some cases, the steps shown or described can be performed in a different order than those described herein, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. Thus, the present invention is not limited to any particular combination of hardware and software.

[0168] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, or improvements made within the principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for generating intelligent scenes, characterized in that, include: When a state change is detected in the first device, a device state change message associated with the first device is obtained, wherein the device state change message is used to indicate the state change that has occurred in the first device; A device list is generated based on device status change messages, wherein the device list includes at least one device identifier, and the device identifier corresponds to a device that is allowed to generate a smart scene; In response to an interactive operation performed on the device list, a target device is selected, and a target intelligent scene is generated based on the target device, the first device, and the state changes that occur to the first device, wherein the target device, the first device, and the state changes that occur to the first device are automatically associated in the target intelligent scene; The step of generating a device list based on device state change messages includes: determining a first tail entity vector based on the device state change messages, wherein the first device corresponds to a first head entity vector, the state change of the first device corresponds to a first relation vector, and the first head entity vector, the first relation vector, and the first tail entity vector together constitute a first triplet in a pre-determined intelligent scene triplet set; generating the device list based on the first tail entity vector, wherein the similarity between the tail entity vectors corresponding to the devices in the device list and the first tail entity vector satisfies a preset condition.

2. The method according to claim 1, characterized in that, Determining the first tail entity vector based on the device status change message includes: Extract the first header entity vector and the first relationship vector based on the device status change message; The first head entity vector and the first relation vector are input into a pre-trained target neural network model to obtain the first tail entity vector. The target neural network model is a model obtained by training an initial neural network model to be trained using a set of sample triples. The set of sample triples includes a set of positive sample triples and a set of negative sample triples. The set of positive sample triples is a set of labeled triples. The set of negative sample triples is a set of triples obtained by replacing the head entity vector or the tail entity vector of the triples in the set of positive sample triples. Specifically, during the iterative training of the initial neural network model, when the input to the initial neural network model is a negative sample triplet from the negative sample triplet set, and the distance between the sum vector formed by the negative sample head entity vector and the negative sample relation vector and the negative sample tail entity vector is greater than a first distance threshold, and / or, when the input to the initial neural network model is a positive sample triplet from the positive sample triplet set, and the distance between the sum vector formed by the positive sample head entity vector and the positive sample relation vector and the positive sample tail entity vector is less than a second distance threshold, it is confirmed that the loss function of the initial neural network model satisfies a preset loss condition, and the initial neural network model is determined as the target neural network model; when the loss function does not satisfy the preset loss condition, the parameters of the initial neural network model are adjusted until the loss function satisfies the preset loss condition.

3. The method according to claim 1, characterized in that, The method further includes: Acquire historical data of device status changes, wherein the historical data of device status changes includes a second device that has undergone a status change, a first time information of the second device undergoing a status change, a third device that has undergone a status change within a preset time interval after the first time information, and a second time information of the third device undergoing a status change. The second triplet in the set of intelligent scene triplets is constructed based on the historical data of device state changes. The second triplet includes the second head entity vector corresponding to the second device, the second relation vector corresponding to the state change of the second device, and the second tail entity vector corresponding to the third device.

4. The method according to claim 3, characterized in that, The method further includes: If the device state change history data does not include the third device and the second time information, a third triplet is constructed in the intelligent scene triplet set based on the device state change history data. The third triplet includes an empty third head entity vector, a third relation vector corresponding to the second time information, and a third tail entity vector corresponding to the second device; or, When the first time information and the second time information included in the device state change history data indicate that the second device and the third device need to keep their state changes synchronized, a fourth triplet is constructed in the intelligent scene triplet set based on the device state change history data. The fourth triplet includes the second head entity vector, a fourth relation vector indicating that the second device and the third device need to keep their state changes synchronized, and the second tail entity vector.

5. The method according to claim 1, characterized in that, The step of generating the device list based on the first tail entity vector includes: Obtain a set of tail entity vectors whose distance values ​​from the first tail entity vector satisfy a preset distance threshold, wherein the sorting position of the tail entity vectors in the set of tail entity vectors is negatively correlated with the distance value; The device list is generated based on the set of tail entity vectors, wherein the devices in the device list are arranged according to the sorting position of the tail entity vectors.

6. The method according to claim 1, characterized in that, The process of generating a device list based on device status change messages includes: Display a target prompt message on the target client, wherein the target prompt message is used to indicate that the first device has undergone a state change and that the generation of a smart scene associated with the first device is currently allowed; In response to the confirmation operation performed on the target prompt message, a list of state changes of the first device is displayed, wherein the list of state changes includes state changes that the first device is allowed to perform; In response to a selection operation performed on the list of state changes, the list of devices is generated based on the selected target state change.

7. A device for generating intelligent scenes, characterized in that, include: The acquisition module, upon detecting a state change in the first device, acquires a device state change message associated with the first device, wherein the device state change message is used to indicate the state change that has occurred in the first device; The generation module generates a device list based on device status change messages, wherein the device list includes at least one device identifier, and the device identifier corresponds to a device that is allowed to generate a smart scene; The processing module is configured to respond to an interactive operation performed on the device list, select a target device, and generate a target intelligent scene based on the target device, the first device, and the state changes of the first device, wherein the target device, the first device, and the state changes of the first device are automatically associated in the target intelligent scene; The device is used to generate a device list based on device state change messages in the following manner: determining a first tail entity vector based on the device state change messages, wherein the first device corresponds to a first head entity vector, the state change of the first device corresponds to a first relation vector, and the first head entity vector, the first relation vector, and the first tail entity vector together constitute a first triplet in a pre-determined intelligent scene triplet set; generating the device list based on the first tail entity vector, wherein the similarity between the tail entity vectors corresponding to the devices in the device list and the first tail entity vector satisfies a preset condition.

8. A system for generating intelligent scenes, characterized in that, include: A detection device is used to acquire a device state change message associated with the first device when a state change is detected in the first device, wherein the device state change message is used to indicate the state change that has occurred in the first device; A server is configured to generate a device list based on device status change messages, wherein the device list includes at least one device identifier, the device identifier corresponding to a device that is allowed to generate a smart scene; An application is configured to select a target device in response to an interactive operation performed on the device list, and generate a target smart scene based on the target device, the first device, and the state changes that occur to the first device, wherein the target device, the first device, and the state changes that occur to the first device are automatically associated in the target smart scene; The step of generating a device list based on device state change messages includes: determining a first tail entity vector based on the device state change messages, wherein the first device corresponds to a first head entity vector, the state change of the first device corresponds to a first relation vector, and the first head entity vector, the first relation vector, and the first tail entity vector together constitute a first triplet in a pre-determined intelligent scene triplet set; generating the device list based on the first tail entity vector, wherein the similarity between the tail entity vectors corresponding to the devices in the device list and the first tail entity vector satisfies a preset condition.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, wherein the computer program, when executed by a processor, implements the steps of the method described in any one of claims 1 to 6.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method described in any one of claims 1 to 6.

Images