A method, apparatus, device, and readable storage medium for generating user profiles.

Abstract

This invention belongs to the field of user profiling technology, and provides a method, apparatus, device, and readable storage medium for generating user profiles. The method includes: acquiring first data, the first data including data information related to users; preprocessing the first data using a decision tree algorithm, and generating sample data based on the preprocessed first data; constructing an initial data model based on a backpropagation (BP) neural network algorithm, training the initial data model using the sample data to obtain user profile feature information, and generating a user profile based on the user profile feature information. This invention enables the model to process large amounts of data through neural network algorithms and improves the model's processing efficiency through self-learning. Furthermore, the model structure constructed using sample data training improves the accuracy of data processing, reduces the workload of staff, and enhances the accuracy of user profile data analysis.

Description

Technical Field

[0001] This invention relates to the field of user profiling technology, and more specifically, to a method, apparatus, device, and readable storage medium for generating user profiles. Background Technology

[0002] The Internet of Things (IoT) is an information carrier based on the Internet, traditional telecommunications networks, etc., that enables all ordinary physical objects that can be independently addressed to form an interconnected network. With social development and progress, the construction of user profiles is becoming increasingly important. User profiles can utilize multi-dimensional views of data to objectively and realistically reflect users' behavioral patterns, habits, and service needs, providing necessary technical support for improving service capabilities and data analysis across various fields.

[0003] The core of current user profiling methods in the technology field is to collect and organize data, and then conduct user profile modeling and analysis. Existing methods suffer from small sample sizes during data collection and processing, which is not conducive to accurately reflecting the objective situation. Furthermore, the data preprocessing methods are too simplistic, hindering sample accuracy. In addition, modeling and analysis often rely on manual labeling, which is not only labor-intensive but also results in low accuracy between the labels and actual preferences, leading to unsatisfactory performance. Summary of the Invention

[0004] The purpose of this invention is to provide a method, apparatus, device, and readable storage medium for generating user profiles, in order to improve the above-mentioned problems.

[0005] To achieve the above objectives, the embodiments of this application provide the following technical solutions:

[0006] On one hand, embodiments of this application provide a method for generating user profiles, the method comprising:

[0007] Obtain first data, which includes data information associated with the user;

[0008] The first data is preprocessed using a decision tree algorithm, and sample data is generated based on the preprocessed first data.

[0009] An initial data model is constructed based on the BP neural network algorithm. The initial data model is trained using the sample data to obtain user profile feature information. A user profile is then generated based on the user profile feature information.

[0010] Secondly, embodiments of this application provide a user profile generation apparatus, the apparatus including an acquisition module, a processing module, and a training module.

[0011] The acquisition module is used to acquire first data, which includes data information related to the user;

[0012] The processing module is used to preprocess the first data using a decision tree algorithm and generate sample data based on the preprocessed first data.

[0013] The training module is used to construct an initial data model based on the BP neural network algorithm, train the initial data model using the sample data to obtain user profile feature information, and generate a user profile based on the user profile feature information.

[0014] Thirdly, embodiments of this application provide a user profile generation device, the device including a memory and a processor. The memory is used to store a computer program; the processor is used to execute the computer program to implement the steps of the user profile generation method described above.

[0015] Fourthly, embodiments of this application provide a readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the above-described user profile generation method.

[0016] The beneficial effects of this invention are as follows:

[0017] This invention enables the acquisition of large amounts of data related to target users through Internet of Things (IoT) technology. The data is then subdivided, categorized, and processed. Furthermore, when building the data model structure, a neural network algorithm is used to enable the model to process large amounts of data and improve its processing efficiency through self-learning. The model structure, trained using sample data, enhances the accuracy of data processing, reduces the workload of staff, and improves the accuracy of user profile data analysis.

[0018] This invention also performs clustering processing on historical user profile feature information. Considering that after clustering, corresponding recommended content needs to be matched for each cluster, and that some clusters may have a small amount of data, staff may encounter inaccurate matching when matching recommended content due to insufficient data, this step processes the data in each cluster to ensure that the number of data in each cluster reaches a preset value, thereby helping staff match more accurate recommended content. After matching corresponding recommended content for each cluster, once the user profile feature information is obtained, recommended content can be quickly matched, improving the user experience.

[0019] Other features and advantages of the invention will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing embodiments of the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description, claims, and drawings. Attached Figure Description

[0020] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a schematic diagram of the user profile generation method described in this embodiment of the invention;

[0022] Figure 2 This is a schematic diagram of the user profile generation device described in an embodiment of the present invention;

[0023] Figure 3 This is a schematic diagram of the user profile generation device described in this embodiment of the invention. Detailed Implementation

[0024] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0025] It should be noted that similar reference numerals or letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this invention, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0026] Example 1

[0027] like Figure 1 As shown in the figure, this embodiment provides a method for generating user profiles, which includes steps S1, S2 and S3.

[0028] Step S1: Obtain first data, which includes data information related to the user;

[0029] In this step, user-related data can be acquired in real time using the Internet of Things (IoT), or it can be obtained from all historical data, or from historical data within a preset historical period (i.e., historical data can be updated by setting a time). User-related data includes various types of information, such as basic user information (phone number, gender, age, etc.), user consumption information (purchased goods information), user entertainment behavior information, and user driving behavior information (charging or refueling information, entry and exit information), etc. This part of the data covers all aspects of the user's life; for example, user entertainment behavior information includes one or more of the following: music preferences, radio preferences, video preferences, and game preferences.

[0030] Step S2: Preprocess the first data using the decision tree algorithm, and generate sample data based on the preprocessed first data;

[0031] The specific implementation steps of this step include step S21 and step S22;

[0032] Step S21: Input the data information into the decision tree algorithm in sequence, classify and process it according to the preset data format to obtain multiple data subsets. The preset data format includes time, location and behavior dimensions. Each data subset represents the user's behavior data at different times and locations.

[0033] In this step, each subset of data obtained can be understood as, for example, a user turning on the lights when they get home from get off work, or a user opening the refrigerator when they get home from get off work;

[0034] Step S22: Standardize each data subset and record each standardized data subset as a sample data.

[0035] In this step, standardization involves editing each data subset to a uniform format (font size, information arrangement, image display method, etc.);

[0036] Step S3: Construct an initial data model based on the BP neural network algorithm, train the initial data model using the sample data to obtain user profile feature information, and generate a user profile based on the user profile feature information.

[0037] The specific implementation steps of this step include step S31 and step S32;

[0038] Step S31: Construct the initial data model, wherein the initial data model includes an input layer, a hidden layer and an output layer, and the hidden layer includes a BP neural network algorithm;

[0039] Step S32: Input the sample data into the initial data model for training. During training, the sample data in the training set is input from the input layer into the hidden layer. The BP neural network algorithm is used to process the sample data in the hidden layer to obtain the feature value corresponding to each sample data. The feature value is output from the output layer and recorded as the user profile feature information.

[0040] After generating the user profile, step S4 is also included;

[0041] Step S4: Sort the multiple daily behavior habit information included in the user profile according to a preset sorting method, and push the sorted information to the user.

[0042] In this step, after constructing the user profile, which contains multiple pieces of information on daily behavioral habits, each piece of information is sorted alphabetically from A to Z and then pushed to the user. After being pushed to the user, the system can also receive user action information for any behavioral habit (e.g., selecting this behavioral habit). In response to the user's selection information, the system helps the user plan and process the selected behavioral habit information in advance (e.g., if the selected behavioral habit is to turn on the lights upon entering the room, the system will automatically turn on the lights when the user enters the room).

[0043] In addition to the steps mentioned above, after obtaining the user profile feature information, the following steps are also included:

[0044] Step S5: Obtain historical user profile feature information, and perform clustering processing on the historical user profile feature information according to the density-based clustering method with noise to obtain multiple clusters, and match the corresponding recommended content for each cluster.

[0045] In this step, the density-based clustering method with noise is the DBSCAN algorithm. The significant advantages of the DBSCAN algorithm are its fast clustering speed and its ability to effectively handle noisy points and discover spatial clusters of arbitrary shapes. In this step, considering that after clustering, each cluster needs to be matched with corresponding recommended content, and also considering that some clusters may have a small number of data points, potentially leading to inaccurate matching when matching recommended content, this step processes the cluster data to ensure that the number of data points in each cluster reaches a preset value, thereby helping staff match more accurate recommended content. The specific implementation steps of this step include steps S51 and S52.

[0046] Step S51: Using a density-based clustering method with noise, the historical user profile feature information is clustered to obtain multiple clusters to be processed. The distance between each cluster to be processed and other clusters to be processed is calculated to obtain multiple first distance calculation results. All the first distance calculation results are summed to obtain a second distance calculation result.

[0047] In this step, in addition to using density-based clustering methods with noise, K-means clustering or the CLARANS algorithm can also be used.

[0048] Step S52: Based on the magnitude of the second distance calculation result, process each cluster to be processed sequentially to ensure that the number of historical user profile feature information contained in each cluster reaches a preset value. The specific implementation steps of this step include step S521;

[0049] Step S521: Starting from the cluster to be processed corresponding to the largest second distance calculation result, traverse the cluster and calculate whether the number of historical user profile feature information contained in each cluster to be processed is less than a preset value. If it is less, add the historical user profile feature information closest to the centroid of the current cluster to the current cluster to be processed until the number of historical user profile feature information in the current cluster to be processed reaches the preset value. If it is greater, calculate the distance between each historical user profile feature information in the current cluster to each centroid except the current centroid, and obtain multiple third distance calculation results. Sum all the third distance calculation results to obtain a fourth distance calculation result. In the current cluster to be processed, remove historical user profile feature information in ascending order according to the size of the fourth distance calculation result corresponding to each historical user profile feature information until the number of historical user profile feature information in the current cluster to be processed reaches the preset value. After the traversal is completed, multiple clusters are obtained.

[0050] Step S6: Calculate the similarity between the user profile feature information and each of the historical user profile feature information, take the historical user profile feature information corresponding to the largest similarity calculation result as the first information, and push the recommended content corresponding to the first information to the user.

[0051] In this step, after clustering and recommended content matching are completed sequentially, and new historical user profile feature information is obtained, similarity calculation can be performed directly to quickly obtain the recommended content for the user. In addition, data can be continuously updated to ensure that the recommended content is more accurate and improve the user experience.

[0052] Example 2

[0053] like Figure 2As shown, this embodiment provides a user profile generation device, which includes an acquisition module 701, a processing module 702, and a training module 703.

[0054] The acquisition module 701 is used to acquire first data, which includes data information related to the user;

[0055] The processing module 702 is used to preprocess the first data using a decision tree algorithm and generate sample data based on the preprocessed first data.

[0056] The training module 703 is used to construct an initial data model based on the BP neural network algorithm, train the initial data model using the sample data to obtain user profile feature information, and generate a user profile based on the user profile feature information.

[0057] In one specific embodiment of this disclosure, the processing module 702 further includes a classification unit 7021 and a processing unit 7022.

[0058] The classification unit 7021 is used to input the data information into the decision tree algorithm in sequence, classify it according to a preset data format, and obtain multiple data subsets. The preset data format includes time, location and behavior dimensions. Each data subset represents the user's behavior data at different times and locations.

[0059] The processing unit 7022 is used to perform standardization processing on each of the data subsets, and to record each of the standardized data subsets as a sample data.

[0060] In one specific embodiment of this disclosure, the training module 703 further includes a construction unit 7031 and a training unit 7032.

[0061] Construction unit 7031 is used to construct the initial data model, wherein the initial data model includes an input layer, a hidden layer and an output layer, and the hidden layer includes a BP neural network algorithm;

[0062] The training unit 7032 is used to input the sample data into the initial data model for training. During training, the sample data in the training set is input from the input layer into the hidden layer. The BP neural network algorithm is used to process the sample data in the hidden layer to obtain the feature value corresponding to each sample data. The feature value is output from the output layer and recorded as the user profile feature information.

[0063] In one specific embodiment of this disclosure, the device further includes a push module 704.

[0064] The push module 704 is used to sort multiple daily behavior habit information included in the user profile according to a preset sorting method, and then push the sorted information to the user.

[0065] It should be noted that the specific manner in which each module performs its operation in the apparatus described in the above embodiments has been described in detail in the embodiments of the method, and will not be elaborated here.

[0066] Example 3

[0067] Corresponding to the above method embodiments, this disclosure also provides a user profile generation device. The user profile generation device described below can be referred to in correspondence with the user profile generation method described above.

[0068] Figure 3 This is a block diagram of a user profile generation device 800 according to an exemplary embodiment. Figure 3 As shown, the user profile generation device 800 may include a processor 801 and a memory 802. The user profile generation device 800 may also include one or more of a multimedia component 803, an I / O interface 804, and a communication component 805.

[0069] The processor 801 controls the overall operation of the user profile generation device 800 to complete all or part of the steps in the user profile generation method described above. The memory 802 stores various types of data to support the operation of the user profile generation device 800. This data may include, for example, instructions for any application or method operating on the user profile generation device 800, and application-related data such as contact data, sent and received messages, images, audio, video, etc. The memory 802 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read-Only Memory (EPROM), Programmable Read-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. Multimedia component 803 may include a screen and an audio component. The screen may be, for example, a touchscreen, and the audio component is used to output and / or input audio signals. For example, the audio component may include a microphone for receiving external audio signals. The received audio signals may be further stored in memory 802 or transmitted via communication component 805. The audio component also includes at least one speaker for outputting audio signals. I / O interface 804 provides an interface between processor 801 and other interface modules, such as a keyboard, mouse, buttons, etc. These buttons may be virtual or physical buttons. Communication component 805 is used for wired or wireless communication between the user profile generation device 800 and other devices. Wireless communication may include Wi-Fi, Bluetooth, Near Field Communication (NFC), 2G, 3G, or 4G, or a combination of these. Therefore, the corresponding communication component 805 may include a Wi-Fi module, a Bluetooth module, or an NFC module.

[0070] In an exemplary embodiment, the user profile generation device 800 may be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the user profile generation method described above.

[0071] In another exemplary embodiment, a computer-readable storage medium including program instructions is also provided, which, when executed by a processor, implement the steps of the user profile generation method described above. For example, the computer-readable storage medium may be the memory 802 including the program instructions, which may be executed by the processor 801 of the user profile generation device 800 to complete the user profile generation method described above.

[0072] Example 4

[0073] Corresponding to the above method embodiments, this disclosure also provides a readable storage medium. The readable storage medium described below can be referred to in conjunction with the user profile generation method described above.

[0074] A readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the user profile generation method described in the above method embodiments.

[0075] The readable storage medium can specifically be a USB flash drive, external hard drive, read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk, or any other readable storage medium capable of storing program code.

[0076] 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, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for generating user profiles, characterized in that, include: Obtain first data, which includes data information associated with the user; The first data is preprocessed using a decision tree algorithm, and sample data is generated based on the preprocessed first data. An initial data model is constructed based on the BP neural network algorithm. The initial data model is trained using the sample data to obtain user profile feature information. A user profile is then generated based on the user profile feature information. Historical user profile feature information is obtained, and a density-based clustering method with noise is used to cluster the historical user profile feature information to obtain multiple clusters to be processed. The distance between each cluster to be processed and other clusters to be processed is calculated to obtain multiple first distance calculation results. All the first distance calculation results are summed to obtain a second distance calculation result. Starting from the cluster corresponding to the largest second distance calculation result, the process iterates through the clusters to be processed. It calculates whether the number of historical user profile features in each cluster is less than a preset value. If less, it adds the historical user profile feature closest to the centroid of the current cluster to the current cluster, until the number of historical user profile features in the current cluster reaches the preset value. If greater, it calculates the distance between each historical user profile feature in the current cluster and every centroid except the current centroid, obtaining multiple third distance calculation results. All third distance calculation results are summed to obtain a fourth distance calculation result. In the current cluster, historical user profile features are removed sequentially from smallest to largest according to the magnitude of the fourth distance calculation result corresponding to each historical user profile feature, until the number of historical user profile features in the current cluster reaches the preset value. After the traversal is complete, multiple clusters are obtained. Match corresponding recommended content to each cluster; Calculate the similarity between the user profile feature information and each of the historical user profile feature information, take the historical user profile feature information corresponding to the largest similarity calculation result as the first information, and push the recommended content corresponding to the first information to the user.

2. The user profile generation method according to claim 1, characterized in that, The step of preprocessing the first data using a decision tree algorithm and generating sample data based on the preprocessed first data includes: The data information is sequentially input into the decision tree algorithm and classified according to a preset data format to obtain multiple data subsets. The preset data format includes time, location, and behavior dimensions. Each data subset represents the user's behavior data at different times and locations. Each of the data subsets is standardized, and each standardized data subset is recorded as a sample data.

3. The user profile generation method according to claim 1, characterized in that, The initial data model is constructed based on the BP neural network algorithm, and the initial data model is trained using the sample data to obtain user profile feature information, including: Construct the initial data model, wherein the initial data model includes an input layer, a hidden layer and an output layer, and the hidden layer includes a BP neural network algorithm; The sample data is input into the initial data model for training. During training, the sample data in the training set is input from the input layer into the hidden layer. The BP neural network algorithm is used to process the sample data in the hidden layer to obtain the feature value corresponding to each sample data. The feature value is output from the output layer and recorded as the user profile feature information.

4. The user profile generation method according to claim 1, characterized in that, After generating the user profile, the following is also included: The user profile includes multiple daily behavioral habit information items, which are sorted according to a preset sorting method, and then pushed to the user.

5. A user profile generation device, characterized in that, include: The acquisition module is used to acquire first data, which includes data information related to the user; The processing module is used to preprocess the first data using a decision tree algorithm and generate sample data based on the preprocessed first data. The training module is used to construct an initial data model based on the BP neural network algorithm, train the initial data model using the sample data to obtain user profile feature information, and generate a user profile based on the user profile feature information. Historical user profile feature information is obtained, and a density-based clustering method with noise is used to cluster the historical user profile feature information to obtain multiple clusters to be processed. The distance between each cluster to be processed and other clusters to be processed is calculated to obtain multiple first distance calculation results. All the first distance calculation results are summed to obtain a second distance calculation result. Starting from the cluster corresponding to the largest second distance calculation result, the process iterates through the clusters to be processed. It calculates whether the number of historical user profile features in each cluster is less than a preset value. If less, it adds the historical user profile feature closest to the centroid of the current cluster to the current cluster, until the number of historical user profile features in the current cluster reaches the preset value. If greater, it calculates the distance between each historical user profile feature in the current cluster and every centroid except the current centroid, obtaining multiple third distance calculation results. All third distance calculation results are summed to obtain a fourth distance calculation result. In the current cluster, historical user profile features are removed sequentially from smallest to largest according to the magnitude of the fourth distance calculation result corresponding to each historical user profile feature, until the number of historical user profile features in the current cluster reaches the preset value. After the traversal is complete, multiple clusters are obtained. Match corresponding recommended content to each cluster; Calculate the similarity between the user profile feature information and each of the historical user profile feature information, take the historical user profile feature information corresponding to the largest similarity calculation result as the first information, and push the recommended content corresponding to the first information to the user.

6. A user profile generation device, characterized in that, include: Memory, used to store computer programs; A processor, configured to implement the steps of the user profile generation method as described in any one of claims 1 to 4 when executing the computer program.

7. A readable storage medium, characterized in that: The readable storage medium stores a computer program that, when executed by a processor, implements the steps of the user profile generation method as described in any one of claims 1 to 4.

Images