CN117290689A

CN117290689A - Smart home-based user binding method and system

Info

Publication number: CN117290689A
Application number: CN202311231851.2A
Authority: CN
Inventors: 彭凡; 王进; 刘阳; 王韬
Original assignee: Hubei Taisheng Technology Co ltd
Current assignee: Hubei Taisheng Technology Co ltd
Priority date: 2023-09-21
Filing date: 2023-09-21
Publication date: 2023-12-26

Abstract

The invention provides a user binding method and system based on smart home, and relates to the technical field of data processing. In the invention, each user to be bound in a plurality of users to be bound is marked as a target user in turn; based on the corresponding historical user behavior data and current user behavior data, performing user characteristic analysis processing on the target user, and analyzing target user characteristic information corresponding to the target user; based on the corresponding target user characteristic information, carrying out classification combination processing on a plurality of users to be bound to form at least one user classification combination, wherein the correlation between the target user characteristic information corresponding to each two users to be bound in different user classification combinations does not meet the correlation condition; based on the user classification combination, user binding processing is carried out on a plurality of users to be bound so as to form user binding relations among the users to be bound. Based on the method, the reliability of user binding can be improved.

Description

Smart home-based user binding method and system

Technical Field

The invention relates to the technical field of data processing, in particular to a user binding method and system based on smart home.

Background

The smart home is also called as an intelligent home, is a living environment in which an intelligent home system is installed by taking a house as a platform, and the process of implementing the intelligent home system is called as intelligent home integration. The residence is taken as a platform, the comprehensive wiring technology, the network communication technology, the safety precaution technology, the automatic control technology and the audio and video technology are utilized to integrate facilities related to the living of the residence, an efficient management system of residence facilities and family schedule matters is constructed, the residence safety, convenience, comfort and artistry are improved, and the environment-friendly and energy-saving living environment is realized. Smart home, smart and home, needs to be understood in two parts. The home is all kinds of equipment for people to live; the intelligent home is an important point that the intelligent home should be highlighted, automatic control and management should be achieved, manual operation control is not needed, the use habit of the current user can be learned, and the requirements of people are met.

In the application of smart home, the relationship binding process may need to be performed on the corresponding user, however, in the prior art, there is a problem that the reliability of the relationship binding of the user is poor.

Disclosure of Invention

Therefore, the invention aims to provide a user binding method and system based on smart home so as to improve the reliability of user binding.

In order to achieve the above purpose, the embodiment of the present invention adopts the following technical scheme:

a user binding method based on smart home comprises the following steps:

marking each user to be bound in a plurality of users to be bound as a target user in turn;

based on the corresponding historical user behavior data and current user behavior data, carrying out user characteristic analysis processing on the target user, and analyzing target user characteristic information corresponding to the target user;

based on the corresponding target user characteristic information, carrying out classification combination processing on the plurality of users to be bound to form at least one user classification combination, wherein the correlation between the target user characteristic information corresponding to each two users to be bound in the same user classification combination meets a preset correlation condition, and the correlation between the target user characteristic information corresponding to each two users to be bound in different user classification combinations does not meet the correlation condition;

and carrying out user binding processing on the plurality of users to be bound based on the user classification combination so as to form a user binding relationship among the users to be bound.

In some preferred embodiments, in the smart home-based user binding method, the step of performing a classification and combination process on the plurality of users to be bound based on the corresponding target user feature information to form at least one user classification and combination includes:

For each two to-be-bound users in the plurality of to-be-bound users, carrying out consistency or correlation analysis on the target user characteristic information corresponding to the two to-be-bound users, placing each two to-be-bound users with consistency or correlation between the corresponding target user characteristic information into the same user classification combination, and placing each two to-be-bound users with non-consistency or correlation between the corresponding target user characteristic information into different user classification combinations to form at least one user classification combination corresponding to the plurality of to-be-bound users, wherein each user classification combination comprises at least one to-be-bound user; or alternatively

And constructing a target user set based on the multiple to-be-bound users, randomly selecting one to-be-bound user from the target user set as a target to-be-bound user, placing the target to-be-bound user into an empty user classification combination, screening the target to-be-bound user from the target user set, sequentially traversing other to-be-bound users in the target user set to determine a new target to-be-bound user from the target user set after each other to-be-bound user is sequentially traversed, and determining the user classification combination corresponding to the target to-be-bound user under the condition that consistency or correlation exists between the characteristic information of the target user corresponding to the other to-be-bound user obtained by current traversing and the characteristic information of the target user corresponding to the target to-be-bound user.

In some preferred embodiments, in the smart home based user binding method, the step of performing user binding processing on the plurality of users to be bound based on the user classification combination to form a user binding relationship between the users to be bound includes:

for each two to-be-bound users belonging to the same user classification combination, carrying out binding processing on the two to-be-bound users so as to form a user binding relationship between the two to-be-bound users;

and for each two users to be bound which do not belong to the same user classification combination, carrying out user binding relation determination processing on the two users to be bound based on at least one other user characteristic information except the target user characteristic information so as to determine whether to establish the corresponding user binding relation between the two users to be bound.

In some preferred embodiments, in the smart home-based user binding method, the step of determining, for each two users to be bound that do not belong to the same user classification combination, a user binding relationship between the two users to be bound based on at least one other user characteristic information other than the target user characteristic information, so as to determine whether to establish a corresponding user binding relationship between the two users to be bound includes:

For any two users to be bound which do not belong to the same user classification combination, marking the two users to be bound as a first user to be bound and a second user to be bound respectively;

determining the user association equipment of the first user to be bound so as to output a first user association equipment set corresponding to the first user to be bound, determining the user association equipment of the second user to be bound so as to output a second user association equipment set corresponding to the second user to be bound, wherein each first user association equipment included in the first user association equipment set is associated with the first user to be bound, and each second user association equipment included in the second user association equipment set is associated with the second user to be bound;

and based on first user associated equipment included in the first user associated equipment set and second user associated equipment included in the second user associated equipment set, determining a user binding relationship between the first user to be bound and the second user to be bound, so as to determine whether to establish the user binding relationship between the first user to be bound and the second user to be bound.

In some preferred embodiments, in the smart home-based user binding method, the step of determining whether to establish a user binding relationship between the first user to be bound and the second user to be bound by performing a user binding relationship determination process on the first user to be bound and the second user to be bound based on a first user associated device included in the first user associated device set and a second user associated device included in the second user associated device set includes:

based on first user associated equipment included in the first user associated equipment set and second user associated equipment included in the second user associated equipment set, determining equipment overlap ratio so as to output the equipment overlap ratio between the first user to be bound and the second user to be bound;

collecting a first associated user behavior sequence corresponding to each first user associated device included in the first user associated device set to form a first associated data set, wherein the first associated user behavior sequence is used for reflecting the use behavior of the first user to be bound on the first user associated device;

Collecting a second associated user behavior sequence corresponding to each second user associated device included in the second user associated device set to form a second associated data set, wherein the second associated user behavior sequence is used for reflecting the use behavior of the second user to be bound on the second user associated devices;

based on the first associated user behavior sequence and the second associated user behavior sequence, determining the correlation degree of the behavior sequences so as to output the corresponding correlation degree of the behavior sequences;

and based on the equipment coincidence degree and the behavior sequence correlation degree, determining the user binding relationship between the first user to be bound and the second user to be bound so as to determine whether to establish the user binding relationship between the first user to be bound and the second user to be bound.

In some preferred embodiments, in the smart home based user binding method, the step of determining a behavior sequence relevance based on the first associated user behavior sequence and the second associated user behavior sequence to output a corresponding behavior sequence relevance includes:

performing feature mining processing on a first associated user behavior sequence corresponding to each first user associated device included in the first user associated device set respectively to output a corresponding first associated feature vector set, wherein each first associated feature vector included in the first associated feature vector set corresponds to one first associated user behavior sequence;

Performing feature mining processing on a second associated user behavior sequence corresponding to each second user associated device included in the second user associated device set respectively to output a corresponding second associated feature vector set, wherein each second associated feature vector included in the second associated feature vector set corresponds to one second associated user behavior sequence;

for each first association feature vector, determining a relevant first association feature vector of the first association feature vector in the first association feature vector set, and carrying out relevant salient feature analysis on the first association feature vector based on the relevant first association feature vector so as to output a first salient feature vector corresponding to the first association feature vector, wherein a device interaction relationship between first user association devices corresponding to first association user behavior sequences corresponding to the relevant first association feature vector and first user association devices corresponding to first association user behavior sequences corresponding to the first association feature vector meets a preconfigured target device interaction relationship;

for each second association feature vector, determining a relevant second association feature vector of the second association feature vector in the second association feature vector set, and carrying out relevant significance feature analysis on the second association feature vector based on the relevant second association feature vector so as to output a second significance feature vector corresponding to the second association feature vector, wherein a device interaction relationship between second user association devices corresponding to second association user behavior sequences corresponding to the relevant second association feature vector and second user association devices corresponding to second association user behavior sequences corresponding to the second association feature vector meets a preconfigured target device interaction relationship;

And determining the corresponding behavior sequence correlation degree based on the vector similarity between the first salient feature vector and the second salient feature vector.

In some preferred embodiments, in the smart home-based user binding method, the step of performing a user feature analysis process on the target user based on the corresponding historical user behavior data and the current user behavior data to analyze target user feature information corresponding to the target user includes:

analyzing behavior representing parameters corresponding to current user behavior data based on current behavior feature vectors corresponding to the current user behavior data and historical user behavior data in a historical user behavior sequence, wherein the current user behavior data and the historical user behavior data in the historical user behavior sequence are generated based on the using behaviors of a target user on intelligent household equipment;

determining behavior correlation characterization data between the current user behavior data and the historical user behavior data;

judging whether to perform data screening operation on the historical user behavior sequence or not based on the behavior correlation characterization data and the behavior representation parameters by utilizing a data comparison analysis network;

And under the condition that the historical user behavior sequence is judged to be subjected to data screening operation, extracting historical user behavior data to be screened from the historical user behavior sequence based on the behavior correlation characterization data by utilizing a data screening analysis network, so as to analyze and process user characteristics of the target user according to the historical user behavior sequence and the current user behavior data which are reserved after screening, and analyzing corresponding target user characteristic information.

The embodiment of the invention also provides a user binding system based on smart home, which comprises the following steps:

the user marking module is used for marking each user to be bound in the plurality of users to be bound as a target user in sequence;

the user characteristic analysis module is used for carrying out user characteristic analysis processing on the target user based on the corresponding historical user behavior data and the current user behavior data and analyzing target user characteristic information corresponding to the target user;

the classification combination processing module is used for carrying out classification combination processing on the plurality of users to be bound based on the corresponding target user characteristic information to form at least one user classification combination, wherein the correlation between the target user characteristic information corresponding to each two users to be bound in the same user classification combination meets a preset correlation condition, and the correlation between the target user characteristic information corresponding to each two users to be bound in different user classification combinations does not meet the correlation condition;

And the user binding processing module is used for carrying out user binding processing on the plurality of users to be bound based on the user classification combination so as to form a user binding relationship among the users to be bound.

In some preferred embodiments, in the smart home-based user binding system, the classification and combination processing module is specifically configured to:

In some preferred embodiments, in the smart home based user binding system, the user binding processing module is specifically configured to:

The user binding method and system based on the smart home provided by the embodiment of the invention can be used for marking each user to be bound in a plurality of users to be bound as a target user in sequence; based on the corresponding historical user behavior data and current user behavior data, performing user characteristic analysis processing on the target user, and analyzing target user characteristic information corresponding to the target user; based on the corresponding target user characteristic information, carrying out classification combination processing on a plurality of users to be bound to form at least one user classification combination, wherein the correlation between the target user characteristic information corresponding to each two users to be bound in different user classification combinations does not meet the correlation condition; based on the user classification combination, user binding processing is carried out on a plurality of users to be bound so as to form user binding relations among the users to be bound. Based on the foregoing steps, since the target user characteristic information corresponding to the user to be bound is determined first, the user can be bound reliably based on the target user characteristic information, and thus the reliability of the user binding can be higher compared with the conventional technical scheme of binding simply according to the positions between users.

In order to make the above objects, features and advantages of the present invention more comprehensible, preferred embodiments accompanied with figures are described in detail below.

Drawings

Fig. 1 is a block diagram of a user binding platform based on smart home according to an embodiment of the present invention.

Fig. 2 is a flowchart illustrating steps involved in a smart home-based user binding method according to an embodiment of the present invention.

Fig. 3 is a schematic diagram of each module included in the smart home-based user binding system according to an embodiment of the present invention.

Detailed Description

For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the described embodiments are only some embodiments of the present invention, but not all embodiments of the present invention. The components of the embodiments of the present invention generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations.

Thus, the following detailed description of the embodiments of the invention, as presented in the figures, is not intended to limit the scope of the invention, as claimed, but is merely representative of selected embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.

As shown in fig. 1, the embodiment of the invention provides a user binding platform based on smart home. The smart home based user binding platform may include a memory and a processor.

In detail, the memory and the processor are electrically connected directly or indirectly to realize transmission or interaction of data. For example, electrical connection may be made to each other via one or more communication buses or signal lines. The memory may store at least one software functional module (computer program) that may exist in the form of software or firmware. The processor may be configured to execute an executable computer program stored in the memory, thereby implementing the smart home-based user binding method provided by the embodiment of the present invention (as described below).

Alternatively, in some embodiments, the Memory may be, but is not limited to, random access Memory (Random Access Memory, RAM), read Only Memory (ROM), programmable Read Only Memory (Programmable Read-Only Memory, PROM), erasable Read Only Memory (Erasable Programmable Read-Only Memory, EPROM), electrically erasable Read Only Memory (Electric Erasable Programmable Read-Only Memory, EEPROM), and the like.

Alternatively, in some embodiments, the processor may be a general purpose processor, including a central processing unit (Central Processing Unit, CPU), a network processor (Network Processor, NP), a System on Chip (SoC), etc.; but also Digital Signal Processors (DSPs), application Specific Integrated Circuits (ASICs), field Programmable Gate Arrays (FPGAs) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

Alternatively, in some embodiments, the smart home based user binding platform may be a server with data processing capabilities.

With reference to fig. 2, the embodiment of the invention further provides a smart home based user binding method, which can be applied to the smart home based user binding platform. The method steps defined by the flow related to the smart home based user binding method can be realized by the smart home based user binding platform. The specific flow shown in fig. 2 will be described in detail.

Step S100, marking each to-be-bound user in the plurality of to-be-bound users as a target user in turn.

In the embodiment of the invention, the smart home-based user binding platform can mark each user to be bound in a plurality of users to be bound as a target user in turn.

Step S200, based on the corresponding historical user behavior data and the current user behavior data, user characteristic analysis processing is carried out on the target user, and target user characteristic information corresponding to the target user is analyzed.

In the embodiment of the invention, the smart home-based user binding platform can analyze the user characteristic of the target user based on the corresponding historical user behavior data and current user behavior data, and analyze the target user characteristic information corresponding to the target user.

And step S300, based on the corresponding target user characteristic information, carrying out classification combination processing on the plurality of users to be bound so as to form at least one user classification combination.

In the embodiment of the invention, the smart home-based user binding platform can perform classification and combination processing on the plurality of users to be bound based on the corresponding target user characteristic information so as to form at least one user classification and combination. The correlation between the target user characteristic information corresponding to each two to-be-bound users in the same user classification combination meets a preset correlation condition, and the correlation between the target user characteristic information corresponding to each two to-be-bound users in different user classification combinations does not meet the correlation condition.

Step S400, based on the user classification combination, performing user binding processing on the plurality of users to be bound to form a user binding relationship between the users to be bound.

In the embodiment of the invention, the smart home-based user binding platform can perform user binding processing on the plurality of users to be bound based on the user classification combination so as to form a user binding relationship among the users to be bound.

Based on the foregoing steps, that is, step S100, step S200, step S300, and step S400, since the target user characteristic information corresponding to the user to be bound is determined first, reliable binding can be performed based on the target user characteristic information, and thus, the reliability of user binding can be higher compared with the conventional technical scheme of binding simply according to the location between users.

Optionally, in some embodiments, the step of analyzing the target user characteristic information corresponding to the target user by performing the user characteristic analysis processing on the target user based on the corresponding historical user behavior data and the current user behavior data may further include the following steps, such as the following step S110, step S120, step S130 and step S140.

Step S110, analyzing behavior representation parameters corresponding to the current user behavior data based on the current behavior feature vector corresponding to the current user behavior data and the historical user behavior data in the historical user behavior sequence.

In the embodiment of the invention, the smart home-based user binding platform can analyze the behavior representing parameters corresponding to the current user behavior data based on the current behavior feature vector corresponding to the current user behavior data and the historical user behavior data in the historical user behavior sequence. The current user behavior data and the historical user behavior data in the historical user behavior sequence are generated based on the using behaviors of the target user on the intelligent household equipment, and can be recorded by text data.

Step S120, determining behavior correlation characterization data between the current user behavior data and the historical user behavior data.

In the embodiment of the invention, the smart home-based user binding platform can determine behavior correlation characterization data between the current user behavior data and the historical user behavior data, similarity between corresponding feature vectors and the like.

Step S130, utilizing a data comparison analysis network, judging whether to conduct data screening operation on the historical user behavior sequence based on the behavior correlation characterization data and the behavior representation parameters.

In the embodiment of the invention, the smart home-based user binding platform can judge whether to perform data screening operation on the historical user behavior sequence based on the behavior correlation characterization data and the behavior representing parameters by utilizing a data comparison analysis network.

And step S140, under the condition that the historical user behavior sequence is judged to need to be subjected to data screening operation, extracting historical user behavior data to be screened out from the historical user behavior sequence based on the behavior correlation characterization data by utilizing a data screening analysis network, so as to carry out user characteristic analysis processing on the target user according to the historical user behavior sequence and the current user behavior data which are reserved after screening out, and analyzing corresponding target user characteristic information.

In the embodiment of the invention, the smart home-based user binding platform can extract the historical user behavior data to be screened from the historical user behavior sequence based on the behavior correlation characterization data by using a data screening analysis network under the condition that the historical user behavior sequence is judged to be required to be subjected to data screening operation, so as to analyze and process the user characteristics of the target user according to the historical user behavior sequence and the current user behavior data which are reserved after screening, and analyze the corresponding target user characteristic information, wherein the target user characteristic information can be used as a classification label, such as 1, 2, 3, 4, 5 and the like without meaning information, and can also be specific characteristic meaning information, such as high-value users, low-value users and the like.

Based on the above, that is, based on the steps S110, S120, S130 and S140, before performing the user feature analysis processing, the historical user behavior data is screened, and the current user behavior data is combined in the screening process, so that the reliability of screening is higher, the validity of the screened historical user behavior data is ensured to be higher, the reliability of the user feature analysis can be improved, and the problem of poor reliability in the prior art can be further improved.

Optionally, in some embodiments, the step of determining behavior correlation characterization data between the current user behavior data and the historical user behavior data may further include the following:

determining local behavior correlation characterization data between the current user behavior data and each of the historical user behavior data, for example, calculating correlation between the local behavior correlation characterization data between the current user behavior data and each of the historical user behavior data, for example, calculating semantic similarity and text similarity, so as to obtain corresponding local behavior correlation characterization data; and analyzing global behavior correlation characterization data between the current user behavior data and the historical user behavior sequence based on the local behavior correlation characterization data, for example, performing mean value calculation on each local behavior correlation characterization data to obtain global behavior correlation characterization data between the current user behavior data and the historical user behavior sequence.

Based on this, the step of determining whether to perform the data filtering operation on the historical user behavior sequence based on the behavior correlation characterization data and the behavior representation parameter by using the data comparison analysis network may further include the following contents:

loading the global behavior correlation characterization data and the behavior representation parameters to a data comparison analysis network, and analyzing global correlation degree values between the current user behavior data and the historical user behavior sequences by using the data comparison analysis network;

judging whether to perform data screening operation on the historical user behavior sequence according to the global correlation degree value, for example, when the global correlation degree value is greater than or equal to a preset global correlation degree value, that is, the current user behavior data can represent the historical user behavior sequence, so that it can be determined that the historical user behavior sequence is subjected to data screening operation to reduce the data volume of subsequent processing; in the case that the global correlation degree value is smaller than the preset global correlation degree value, that is, the current user behavior data may not represent the historical user behavior sequence, it may be determined that the data filtering operation is not performed on the historical user behavior sequence.

Optionally, in some embodiments, the smart home based user binding method may further include the following:

based on the first dimension learning cost index of the nth round, carrying out optimization processing on the data comparison analysis network, and judging whether to carry out data screening operation of the nth round or not based on the optimized data comparison analysis network;

under the condition that the data screening operation of the nth round is judged to be needed, analyzing an nth behavior representing parameter corresponding to the current user behavior data based on a historical user behavior sequence which is reserved after the nth round is screened and obtained after the data screening operation of the nth round;

and analyzing a first dimension learning cost index of a y-th round of the data comparison analysis network based on the nth behavior representing parameter and the mth behavior representing parameter (difference between the nth behavior representing parameter and the mth behavior representing parameter), and optimizing the data comparison analysis network based on the first dimension learning cost index of the y-th round to form a new optimized data comparison analysis network, wherein n-m=1.

Optionally, in some embodiments, the step of loading the global behavior correlation characterization data and the behavior representation parameter to be loaded into a data comparison analysis network, and using the data comparison analysis network to analyze the global correlation degree value between the current user behavior data and the historical user behavior sequence may further include the following contents:

Determining nth global behavior correlation characterization data between the current user behavior data and historical user behavior data in a historical user behavior sequence reserved after nth screening;

analyzing a corresponding nth condition feature vector based on the nth behavior representation parameter and the nth global behavior correlation characterization data, and illustratively, encoding the nth behavior representation parameter and the nth global behavior correlation characterization data to obtain a corresponding nth condition feature vector;

according to the optimized data comparison analysis network corresponding to the first dimension learning cost index of the nth round, performing full connection operation on the nth condition feature vector to output a corresponding nth full connection feature vector, for example, weighting the nth condition feature vector based on a first parameter, and then performing superposition processing on the result of the weighting processing based on a second parameter to output a corresponding nth full connection feature vector, wherein the first parameter and the second parameter can be used as network parameters of the data comparison analysis network to optimize;

the nth full-connection feature vector is analyzed according to a weight characterization parameter included in the data comparison analysis network to output a global correlation degree value between the current user behavior data and the historical user behavior sequence, and illustratively, the nth full-connection feature vector may be weighted based on a third parameter, then, an excitation mapping process may be performed on a weighted result, then, a weighted summation process may be performed on the weighted summation process and a negative correlation parameter of the weighted summation process, a sum value between the negative correlation parameter of the weighted summation process and the weighted summation process may be equal to 1, and a sum value between the weighted summation process and the negative correlation parameter of the excitation mapping process may be equal to 1.

local behavior correlation characterization data between the current user behavior data and each of the historical user behavior data is determined, as described in the foregoing.

Based on this, when it is determined that the historical user behavior sequence needs to be subjected to the data filtering operation, the step of extracting, by using a data filtering analysis network, historical user behavior data to be filtered out from the historical user behavior sequence based on the behavior correlation characterization data, so as to perform user feature analysis processing on the target user according to the historical user behavior sequence and the current user behavior data retained after filtering out, and analyze corresponding target user feature information may further include the following contents:

under the condition that the historical user behavior sequence is judged to need to be subjected to data screening operation, loading the x local behavior correlation characterization data corresponding to the x historical user behavior data to be loaded into a data screening analysis network, and analyzing the local correlation degree value between the current user behavior data and the x historical user behavior data by utilizing the data screening analysis network; and judging whether the xth historical user behavior data belongs to historical user behavior data to be screened according to the local correlation degree value, for example, judging that the xth historical user behavior data needs to be screened when the local correlation degree value is larger than or equal to a preset value, and judging that the xth historical user behavior data does not need to be screened when the local correlation degree value is smaller than the preset value; and analyzing the user characteristic of the target user according to the historical user behavior sequence and the current user behavior data which are reserved after screening, and analyzing the corresponding target user characteristic information.

optimizing the data screening analysis network based on a second dimension learning cost index of the a round, and analyzing historical user behavior data to be screened out based on the optimized data screening analysis network;

analyzing an a-th behavior representative parameter of the current user behavior data based on a historical user behavior sequence reserved after a-th screening after a-th data screening operation;

analyzing a first local learning cost index based on (the difference between) the a-th behavior representing parameter and the b-th behavior representing parameter, wherein the b-th behavior representing parameter is obtained based on a historical user behavior sequence reserved after b-th screening after b-th round of data screening operation, and a-b=1;

analyzing the data matching degree, such as semantic relativity, between each historical user behavior data and the current user behavior data in the historical user behavior sequence reserved after the a-th screening, and analyzing a second local learning cost index (the second local learning cost index can have a corresponding relationship with negative relativity between the data matching degree) based on a plurality of the data matching degrees;

And analyzing the second dimension learning cost index of the c-th round of the data screening analysis network based on the first and second local learning cost indexes (such as sum or weighted sum calculation is performed on the first and second local learning cost indexes), and optimizing the data screening analysis network based on the second dimension learning cost index of the c-th round to form an optimized data screening analysis network, wherein c-a=1.

Optionally, in some embodiments, in a case where it is determined that the data filtering operation needs to be performed on the historical user behavior sequence, loading the x-th local behavior correlation characterization data corresponding to the x-th historical user behavior data into a data filtering analysis network, and using the data filtering analysis network to analyze the local correlation degree value between the current user behavior data and the x-th historical user behavior data, the method may further include the following steps:

under the condition that the historical user behavior sequence is judged to need to be subjected to data screening operation, determining the x local behavior correlation characterization data between the current user behavior data and the x historical user behavior data;

Analyzing a corresponding x-th condition feature vector based on the x-th local behavior correlation characterization data, for example, encoding the x-th local behavior correlation characterization data;

according to the optimized data screening analysis network corresponding to the second dimension learning cost index of the a-th round, carrying out full connection operation on the x-th condition feature vector, and outputting a corresponding x-th full connection feature vector as described in the previous step;

and analyzing the x-th full-connection feature vector according to the weight characterization parameters of the data screening analysis network, and outputting a local correlation degree value between the current user behavior data and the x-th historical user behavior data according to the previous correlation description.

Optionally, in some embodiments, the step of analyzing the behavior representing parameter corresponding to the current user behavior data based on the current behavior feature vector corresponding to the current user behavior data and the historical user behavior data in the historical user behavior sequence may further include the following contents:

mining corresponding historical behavior feature vectors based on historical user behavior data in the historical user behavior sequence, such as feature mining or key information mining;

According to the historical behavior feature vector and the current behavior feature vector corresponding to the current user behavior data, analyzing the behavior representing parameter corresponding to the current user behavior data, and performing vector similarity calculation on the historical behavior feature vector and the current behavior feature vector to obtain the behavior representing parameter corresponding to the current user behavior data.

Optionally, in some embodiments, the step of mining out the corresponding historical behavior feature vector based on the historical user behavior data in the historical user behavior sequence may further include the following contents:

and carrying out mean value calculation on the historical behavior feature vector corresponding to the historical user behavior data in the historical user behavior sequence reserved after the n-th screening operation of the data of the n-th round so as to output a corresponding n-th historical behavior feature vector, wherein the n-th historical behavior feature vector is used for analyzing the n-th behavior representative parameter corresponding to the current user behavior data.

Optionally, in some embodiments, the step of performing a classification and combination process on the multiple users to be bound based on the corresponding target user feature information to form at least one user classification and combination may further include the following contents:

For each two to-be-bound users in the plurality of to-be-bound users, carrying out consistency or correlation analysis on the target user characteristic information corresponding to the two to-be-bound users, placing each two to-be-bound users with consistency or correlation between the corresponding target user characteristic information into the same user classification combination, and placing each two to-be-bound users with non-consistency or correlation between the corresponding target user characteristic information into different user classification combinations to form at least one user classification combination corresponding to the plurality of to-be-bound users, wherein each user classification combination comprises at least one to-be-bound user.

and constructing a target user set based on the multiple to-be-bound users, randomly selecting one to-be-bound user from the target user set as a target to-be-bound user, placing the target to-be-bound user into an empty user classification combination, screening the target to-be-bound user from the target user set, sequentially traversing other to-be-bound users in the target user set to determine a user classification combination corresponding to the target to-be-bound user under the condition that consistency or correlation exists between the characteristic information of the target user corresponding to the other to-be-bound user obtained by current traversing and the characteristic information of the target user corresponding to the target to-be-bound user, screening the other to-be-bound user from the target user set until each other to-be-bound user is sequentially traversed, and re-determining one to-be-bound user as a new target to-be-bound user in the target user set to determine the user classification corresponding to the target to-be-bound user, thereby forming at least one circulation combination.

Optionally, in some embodiments, the step of performing a user binding process on the plurality of users to be bound based on the user classification combination to form a user binding relationship between the users to be bound may further include the following contents:

Optionally, in some embodiments, the step of determining, for each two to-be-bound users that do not belong to the same user classification combination, a user binding relationship between the two to-be-bound users based on at least one other user characteristic information other than the target user characteristic information, so as to determine whether to establish a corresponding user binding relationship between the two to-be-bound users may further include:

determining the user association equipment of the first user to be bound so as to output a first user association equipment set corresponding to the first user to be bound, determining the user association equipment of the second user to be bound so as to output a second user association equipment set corresponding to the second user to be bound, wherein each first user association equipment included in the first user association equipment set is associated with the first user to be bound, namely, the first user to be bound uses the first user association equipment, and each second user association equipment included in the second user association equipment set is associated with the second user to be bound, namely, the second user to be bound uses the second user association equipment;

Optionally, in some embodiments, the step of determining whether to establish a user binding relationship between the first to-be-bound user and the second to-be-bound user by performing a user binding relationship determination process on the first to-be-bound user and the second to-be-bound user based on a first user association device included in the first user association device set and a second user association device included in the second user association device set may further include the following contents:

determining the equipment overlap ratio based on first user associated equipment included in the first user associated equipment set and second user associated equipment included in the second user associated equipment set, so as to output the equipment overlap ratio between the first user to be bound and the second user to be bound, wherein the equipment overlap ratio is used for reflecting the number proportion of the same equipment between the sets;

based on the device overlap ratio and the behavior sequence correlation, determining a user binding relationship between the first user to be bound and the second user to be bound to determine whether to establish the user binding relationship between the first user to be bound and the second user to be bound, for example, the device overlap ratio and the behavior sequence correlation may be weighted and summed to obtain a corresponding weighted and summed value, and then, if the weighted and summed value is greater than or equal to a preset value, the user binding relationship may be established between the first user to be bound and the second user to be bound, and if the weighted and summed value is less than the preset value, the user binding relationship may not be established between the first user to be bound and the second user to be bound.

Optionally, in some embodiments, the step of determining a behavior sequence relatedness based on the first associated user behavior sequence and the second associated user behavior sequence to output a corresponding behavior sequence relatedness may further include the following:

for each first association feature vector, determining a relevant first association feature vector of the first association feature vector in the first association feature vector set, and carrying out relevant salient feature analysis on the first association feature vector based on the relevant first association feature vector (illustratively, transposition processing can be carried out on the relevant first association feature vector to obtain a first transposition feature vector, then multiplication can be carried out on the first transposition feature vector and the first association feature vector, vector dimension of the first association feature vector is divided, then activation processing is carried out on a calculation result, finally multiplication can be carried out on the relevant first association feature vector and an activation processing result to realize relevant salient feature analysis, namely, based on the relevant first association feature vector, important and key information is mined from the first association feature vector to output a first salient feature vector corresponding to the first association feature vector, and first user association equipment corresponding to the relevant first association feature vector and first user interaction relation equipment corresponding to the first association feature vector, such as first user interaction relation configuration interaction relation between the first user interaction relation equipment corresponding to the first association feature vector and the first user interaction relation equipment corresponding to the first user interaction relation configuration interaction relation of the first user equipment is satisfied;

For each second association feature vector, determining a relevant second association feature vector of the second association feature vector in the second association feature vector set, and performing relevant salient feature analysis on the second association feature vector based on the relevant second association feature vector to output a second salient feature vector corresponding to the second association feature vector, wherein a device interaction relationship between second user association devices corresponding to second association user behavior sequences corresponding to the relevant second association feature vector and second user association devices corresponding to second association user behavior sequences corresponding to the second association feature vector meets a preset target device interaction relationship, such as maximum interaction data amount;

based on the vector similarity between the first salient feature vector and the second salient feature vector, a corresponding behavior sequence correlation is determined, and the vector similarity between the first salient feature vector and the second salient feature vector may be directly used as the behavior sequence correlation, or may be determined based on other relationships.

With reference to fig. 3, the embodiment of the invention further provides a smart home based user binding system, which can be applied to the smart home based user binding platform. The smart home-based user binding system may include the following software functional modules:

Optionally, in some embodiments, the classification combination processing module is specifically configured to:

Optionally, in some embodiments, the user binding processing module is specifically configured to:

In summary, according to the smart home-based user binding method and system provided by the invention, each user to be bound in a plurality of users to be bound can be marked as a target user in turn; based on the corresponding historical user behavior data and current user behavior data, performing user characteristic analysis processing on the target user, and analyzing target user characteristic information corresponding to the target user; based on the corresponding target user characteristic information, carrying out classification combination processing on a plurality of users to be bound to form at least one user classification combination, wherein the correlation between the target user characteristic information corresponding to each two users to be bound in different user classification combinations does not meet the correlation condition; based on the user classification combination, user binding processing is carried out on a plurality of users to be bound so as to form user binding relations among the users to be bound. Based on the foregoing steps, since the target user characteristic information corresponding to the user to be bound is determined first, the user can be bound reliably based on the target user characteristic information, and thus the reliability of the user binding can be higher compared with the conventional technical scheme of binding simply according to the positions between users.

The above description is only of the preferred embodiments of the present invention and is not intended to limit the present invention, but various modifications and variations can be made to the present invention by those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. The user binding method based on the smart home is characterized by comprising the following steps of:

2. The smart home based user binding method of claim 1, wherein the step of performing a classification and combination process on the plurality of users to be bound based on the corresponding target user characteristic information to form at least one user classification and combination comprises:

3. The smart home based user binding method of claim 1, wherein the step of performing user binding processing on the plurality of users to be bound based on the user classification combination to form a user binding relationship between the users to be bound comprises:

4. The smart home-based user binding method of claim 3, wherein the step of determining, for each two users to be bound that do not belong to the same user classification combination, a user binding relationship between the two users to be bound based on at least one other user characteristic information other than the target user characteristic information, to determine whether to establish a corresponding user binding relationship between the two users to be bound, includes:

5. The smart home based user binding method of claim 4, wherein the step of determining whether to establish a user binding relationship between the first user to be bound and the second user to be bound by performing a user binding relationship determination process on the first user to be bound and the second user to be bound based on a first user associated device included in the first set of user associated devices and a second user associated device included in the second set of user associated devices comprises:

6. The smart home based user binding method of claim 5, wherein the step of determining a behavioral sequence relevance based on the first and second associated user behavior sequences to output a corresponding behavioral sequence relevance comprises:

7. The smart home based user binding method of any one of claims 1-6, wherein the step of performing a user feature analysis process on the target user based on the corresponding historical user behavior data and current user behavior data to analyze target user feature information corresponding to the target user comprises:

8. A smart home based user binding system, comprising:

9. The smart home based user binding system of claim 8, wherein the classification and combination processing module is specifically configured to:

10. The smart home based user binding system of claim 8, wherein the user binding processing module is specifically configured to: