CN115689626B

CN115689626B - User attribute determining method of terminal equipment and electronic equipment

Info

Publication number: CN115689626B
Application number: CN202211348747.7A
Authority: CN
Inventors: 章心宇
Original assignee: Honor Device Co Ltd
Current assignee: Honor Device Co Ltd
Priority date: 2022-10-31
Filing date: 2022-10-31
Publication date: 2024-03-01
Anticipated expiration: 2042-10-31
Also published as: CN115689626A

Abstract

The embodiment of the application provides a method for determining user attributes of terminal equipment and electronic equipment, relates to the field of data processing, and is capable of determining the user attributes of the terminal equipment according to basic parameters and usage data of the terminal equipment and high in accuracy. The method comprises the step of acquiring first data of the first terminal equipment. The first data of the first terminal device is input into the attribute determination model. The attribute determination model is generated by training a preset model by using sample data in advance. The sample data includes first data of a plurality of terminal devices and user attributes. The preset model is a classification model. And determining the user attribute of the first terminal equipment according to the output result of the attribute determination model.

Description

User attribute determining method of terminal equipment and electronic equipment

Technical Field

The embodiment of the application relates to the field of data processing, in particular to a user attribute determining method of terminal equipment and electronic equipment.

Background

With the popularization of terminal devices, research on user attributes of terminal devices is also increasing. The user attributes of the terminal device include the gender of the user of the terminal device, the age of the user, interest preferences, consumption preferences, etc. It can be understood that the user attribute of the terminal device has high value in the aspects of statistics of user distribution of the terminal device, accurate marketing and the like. For example, a product provider whose primary marketing goal is a population of females in users of a brand of terminal device. Marketing costs may be greatly reduced if the product provider may first delineate the target group by user attributes of the branded end devices.

The user attribute of the terminal equipment can be input and stored in the terminal equipment by the user, so that the product provider can acquire the use after the user agrees. However, not all users will input user attributes into the terminal device, so that many terminal devices do not store user attributes, nor are the product providers able to obtain the user attributes of the terminal devices.

Therefore, how to determine the user attribute of the terminal device becomes a problem to be solved.

Disclosure of Invention

The embodiment of the application provides a method for determining the user attribute of terminal equipment and electronic equipment, which can determine the user attribute of the terminal equipment according to the basic parameters and the use data of the terminal equipment and have higher accuracy.

In order to achieve the above purpose, the following technical solutions are adopted in the embodiments of the present application.

In a first aspect, a method for determining a user attribute of a first terminal device is provided, for determining a user attribute of the first terminal device according to first data of the first terminal device. The first data includes basic parameters of the corresponding terminal device and usage data generated after the corresponding terminal device is activated and used. The user attributes include at least one of: user gender, user age group. The method comprises the following steps: first data of a first terminal device are acquired. The first data of the first terminal device is input into the attribute determination model. The attribute determination model is generated by training a preset model by using sample data in advance. The sample data includes first data of a plurality of terminal devices and user attributes. The preset model is a classification model. And determining the user attribute of the first terminal equipment according to the output result of the attribute determination model.

Based on the scheme, the machine learning model trained by the sample data is used, so that the user attribute of the terminal equipment can still be determined according to the basic parameters and the use data of the terminal equipment under the condition that the terminal equipment does not comprise the user attribute, and the accuracy is higher.

In one possible design, before the first data input attribute of the first terminal device is modeled, the method further comprises: and respectively carrying out feature derivation on basic parameters in the first data of the first terminal equipment, and clustering the use data in the first data of the first terminal equipment to obtain feature data of the first terminal equipment. Inputting first data of a first terminal device into a property determination model, comprising: the characteristic data of the first terminal device is input into the attribute determination model. The sample data includes characteristic data of a plurality of terminal devices and user attributes. The characteristic data of the terminal equipment are obtained by carrying out characteristic derivation on basic parameters in the first data of each terminal equipment and clustering the use data in the first data of each terminal equipment. Based on the scheme, feature derivation and feature extraction are carried out on the first data to generate more representative feature data, so that the accuracy of the determined user attribute can be effectively improved.

In one possible design, the first data further includes an activation duration of the corresponding terminal device. Before acquiring the first data of the first terminal device, the method further includes: sample data is acquired. A first sample in the sample data is acquired. The activation time length of each terminal device corresponding to the first sample is longer than the first time length. Training a first model in the preset models by using the first samples to generate a second model in the attribute determination models. The first model is a tree model classifier or a linear classification model. Based on this scheme, training of the first sample in the sample data may be employed during the model training phase. And in the model use stage, when the activation time length of the first terminal equipment is the same as the activation time length of the terminal equipment corresponding to the first sample, inputting the characteristic data of the first terminal equipment into the model trained by the first sample so as to obtain the user attribute of the first terminal equipment. In this way, the accuracy and efficiency of determining the user attributes may be improved.

In one possible design, the first data input attribute determination model of the first terminal device includes: and when the activation time of the first terminal equipment is longer than the first time, inputting the characteristic data of the first terminal equipment into the second model. Determining the user attribute of the first terminal device according to the output result of the attribute determination model, including: and determining the user attribute of the first terminal equipment according to the output result of the second model. Based on this scheme, training of the first sample in the sample data may be employed during the model training phase. And in the model use stage, when the activation time length of the first terminal equipment is the same as the activation time length of the terminal equipment corresponding to the first sample, inputting the characteristic data of the first terminal equipment into the model trained by the first sample so as to obtain the user attribute of the first terminal equipment. In this way, the accuracy and efficiency of determining the user attributes may be improved.

In one possible design, the method further includes, after acquiring the sample data. A second sample in the sample data is acquired. And the activation time length of each terminal device corresponding to the second sample is smaller than or equal to the first time length. And inputting the second sample into the tree model classifier for multiple iterations to obtain leaf node data of each weak classifier in the tree model classifier. The leaf node data is a feature combination of the feature data in the second sample. And carrying out one-bit effective coding on each leaf node data to obtain coded data. And combining the coded data with each characteristic data in the second sample to obtain a third sample. Training a third model in the preset models by using the third sample, and generating a fourth model in the attribute determination model. The fourth model is a linear classification model. Based on this scheme, training of the second sample in the sample data may be employed during the model training phase. And in the model use stage, when the activation time length of the first terminal equipment is the same as the activation time length of the terminal equipment corresponding to the second sample, inputting the characteristic data of the first terminal equipment into the model trained by the second sample so as to obtain the user attribute of the first terminal equipment. In this way, the accuracy and efficiency of determining the user attributes may be improved.

In one possible design, the first data input attribute determination model of the first terminal device includes: and when the activation time length of the first terminal equipment is smaller than or equal to the first time length, inputting the characteristic data of the first terminal equipment into a tree model classifier to obtain a characteristic classification result. And carrying out one-bit effective coding on the feature classification result to obtain a feature coding result. And combining the feature coding result with the feature data of the first terminal equipment to obtain a feature combining result. And taking the sequence of the coded data and the sequence of the feature data in the third sample as references, and carrying out feature alignment on the feature merging result to obtain a feature alignment result. And inputting the characteristic alignment result into a fourth model. Determining the user attribute of the first terminal device according to the output result of the attribute determination model, including: and determining the user attribute of the first terminal equipment according to the output result of the fourth model. Based on this scheme, training of the second sample in the sample data may be employed during the model training phase. And in the model use stage, when the activation time length of the first terminal equipment is the same as the activation time length of the terminal equipment corresponding to the second sample, inputting the characteristic data of the first terminal equipment into the model trained by the second sample so as to obtain the user attribute of the first terminal equipment. In this way, the accuracy and efficiency of determining the user attributes may be improved.

In one possible design, the tree model classifier is a gradient-lifting decision tree.

In one possible design, when the base parameter in the first data of the first terminal device includes a time period in which the terminal device is used daily, performing feature derivation on the base parameter in the first data of the first terminal device includes: the earliest time and the latest time of the daily use of the terminal device are determined according to the daily use time period of the terminal device.

In one possible design, when the usage data in the first data of the first terminal device includes application installation information, clustering the usage data in the first data of the first terminal device includes: and classifying the applications with the installation quantity larger than the first threshold value in the application market to obtain a plurality of application categories. And determining the application category included in the first terminal equipment according to the application installation information.

In one possible design, the basic parameters of the terminal device include at least one of the following: color, model, system version.

In one possible design, the usage data of the terminal device includes at least one of: application installation information, application use information, use time period, camera use frequency, camera setting parameters and system setting parameters.

In a second aspect, an electronic device is provided that includes one or more processors and one or more memories. One or more memories are coupled to the one or more processors, the one or more memories storing computer instructions. The computer instructions, when executed by one or more processors, cause the electronic device to perform the method of determining a user attribute of a terminal device as in any of the first aspects.

In a third aspect, a computer readable storage medium is provided, the computer readable storage medium comprising computer instructions which, when run, perform the method of determining a user attribute of a terminal device as in any of the first aspects.

In a fourth aspect, there is provided a computer program product comprising instructions which, when run on a computer, cause the computer to perform the method of determining a user attribute of a terminal device as in any of the first aspects, in accordance with the instructions.

In a fifth aspect, a system on a chip is provided, the chip comprising processing circuitry and an interface. The processing circuit is configured to call from the storage medium and execute the computer program stored in the storage medium to perform the user attribute determining method of the terminal device as in any one of the second aspects.

It should be appreciated that the technical features of the technical solutions provided in the second aspect, the third aspect, the fourth aspect and the fifth aspect may all correspond to the method for determining the user attribute of the terminal device provided in the first aspect and the possible designs thereof, so that the beneficial effects can be achieved similarly, and are not repeated herein.

Drawings

Fig. 1 is a schematic structural diagram of a terminal device provided in an embodiment of the present application;

fig. 2 is a flowchart of a method for determining a user attribute of a terminal device according to an embodiment of the present application;

FIG. 3 is a flowchart of a model training method according to an embodiment of the present application;

FIG. 4 is a flowchart of yet another model training method provided in an embodiment of the present application;

fig. 5 is a flowchart of a method for determining a user attribute of another terminal device according to an embodiment of the present application;

FIG. 6 is a flowchart of yet another model training method provided in an embodiment of the present application;

FIG. 7 is a schematic diagram of a tree model classifier according to an embodiment of the present application;

fig. 8 is a flowchart of a method for determining a user attribute of another terminal device according to an embodiment of the present application;

fig. 9 is a schematic diagram of the composition of an electronic device according to an embodiment of the present application;

Fig. 10 is a schematic diagram of a system on chip according to an embodiment of the present application.

Detailed Description

The terms "first," "second," and "third," etc. in the embodiments of the present application are used for distinguishing between different objects and not for defining a particular order. Furthermore, the words "exemplary" or "such as" are used to mean serving as an example, instance, or illustration. Any embodiment or design described herein as "exemplary" or "for example" should not be construed as preferred or advantageous over other embodiments or designs. Rather, the use of words such as "exemplary" or "such as" is intended to present related concepts in a concrete fashion.

In order to facilitate understanding of the embodiments of the present application, the following description first refers to the background of the embodiments of the present application.

It should be noted that, in the embodiment of the present application, the user attribute is used to collectively refer to the gender, age group, interest preference, consumption preference, and the like of the user of the terminal device. In some other embodiments, the user attribute may also be referred to as a tag in the user portrait, which will not be described in detail later.

It is easy to understand that the user attribute has high value in the aspects of statistics of user distribution of the terminal equipment, product marketing and the like. For example, the user gender distribution of the terminal device may provide a very important reference for the manufacturer to determine whether a subsequent marketing strategy is biased towards men or women, and the user age bracket distribution of the terminal product may provide an important reference for the manufacturer to determine whether a subsequent marketing strategy is biased towards young or middle-aged.

The users of some terminal devices can input the information of the gender, age and the like of the users into the terminal devices by themselves. Therefore, the product provider can conveniently obtain the user attribute on the premise of user consent. However, more users do not enter user attributes into the terminal device from privacy, data security, etc. concerns. Thus, the product provider does not have a way to obtain the user attribute of the terminal device, so that many business scenarios cannot be expanded.

In order to solve the above problems, the embodiments of the present application provide a method for determining a user attribute of a terminal device, which can determine the user attribute of the terminal device according to a basic parameter and usage data of the terminal device, and has higher accuracy.

The basic parameters of the terminal equipment at least comprise one of the following: color, model, system version. It should be noted that the basic parameters of the terminal device are already stored in the terminal device when shipped from the factory. In other words, the information of color, model, system version, etc. is already stored in the terminal device before the terminal device is activated.

In this embodiment, the color of the terminal device may refer to the name of its manufacturer, for example, the color of a mobile phone of a certain brand may include bright black, white glaze, blue porcelain, flowing gold, etc. The model of the terminal device may also refer to manufacturer's names, such as brand name +x40, brand name +x50, etc. The system version of the terminal device may refer to the version of the operating system installed in the terminal device, such as Version number, of>Version number of the third party operating system provided by the manufacturer, etc. It should be noted that the foregoing is illustrative only and is not meant to limit the present application.

The usage data of the terminal device at least comprises one of the following: application installation information, application use information, use time period, camera use frequency, camera setting parameters and system setting parameters. The usage data of the terminal device refers to data generated after the terminal device is activated and used.

The application installation information may include the number of applications installed in the terminal device, application names, and the like. The application use information may include the number of application uses, the duration of application use, and the like. The usage period may refer to a period in which the terminal device is used. The camera usage frequency may refer to a frequency at which a camera of the terminal device operates. The camera setting parameters may include a focal length parameter, a shutter parameter, an aperture parameter, a sensitivity parameter, and the like. The system setting parameters may include screen refresh rate, wallpaper, font, etc. It should be understood that the above description of usage data is also exemplary only and is not intended to limit the present application thereto.

The application is used for determining the user attribute of the terminal equipment, and the terminal equipment can be a portable terminal, such as a mobile phone, a tablet computer, a wearable device (such as a smart watch), a vehicle-mounted device and the like. Exemplary embodiments of the portable terminal include, but are not limited to, piggy-backOr other operating system. The portable terminal may also be a portable terminal such as a Laptop computer (Laptop) having a touch sensitive surface, e.g. a touch panel. As an example, please refer to fig. 1, which is a schematic structural diagram of a terminal device provided in an embodiment of the present application. The method for determining the user attribute of the terminal device provided in the embodiment of the present application can be applied to the terminal device 100 shown in fig. 1.

As shown in fig. 1, the terminal device 100 may include a processor 101, a display 103, a communication module 102, and the like.

The processor 101 may include one or more processing units, for example: the processor 101 may include an application processor (application processor, AP), a modem processor, a graphics processor (graphics processing unit, GPU), an image signal processor (image signal processor, ISP), a controller, a memory, a video stream codec, a digital signal processor (digital signal processor, DSP), a baseband processor, and/or a neural network processor (neural-network processing unit, NPU), etc. Wherein the different processing units may be separate devices or may be integrated in one or more processors 101.

The controller may be a neural and command center of the terminal device 100. The controller can generate operation control signals according to the instruction operation codes and the time sequence signals to finish the control of instruction fetching and instruction execution.

A memory may also be provided in the processor 101 for storing instructions and data. In some embodiments, the memory in the processor 101 is a cache memory. The memory may hold instructions or data that has just been used or recycled by the processor 101. If the processor 101 needs to reuse the instruction or data, it may be called directly from the memory. Repeated accesses are avoided and the latency of the processor 101 is reduced, thus improving the efficiency of the system.

In some embodiments, the processor 101 may include one or more interfaces. The interfaces may include an integrated circuit (inter-integrated circuit, I2C) interface, an integrated circuit built-in audio (inter-integrated circuit sound, I2S) interface, a pulse code modulation (pulse code modulation, PCM) interface, a universal asynchronous receiver transmitter (universal asynchronous receiver/transmitter, UART) interface, a mobile industry processor interface (mobile industry processor interface, MIPI), a general-purpose input/output (GPIO) interface, a subscriber identity module (subscriber identity module, SIM) interface, and/or a universal serial bus (universal serial bus, USB) interface 111, among others.

The terminal device 100 realizes a display function by a GPU, a display screen 103, an application processor, and the like. The GPU is a microprocessor for image processing, and is connected to the display screen 103 and the application processor. The GPU is used to perform mathematical and geometric calculations for graphics rendering. Processor 101 may include one or more GPUs that execute program instructions to generate or change display information.

The display 103 is used to display images, video streams, and the like.

The communication module 102 may include an antenna 1, an antenna 2, a mobile communication module 102A, and/or a wireless communication module 102B. Taking the communication module 102 as an example, the antenna 1, the antenna 2, the mobile communication module 102A and the wireless communication module 102B are included at the same time.

The wireless communication function of the terminal device 100 can be realized by the antenna 1, the antenna 2, the mobile communication module 102A, the wireless communication module 102B, a modem processor, a baseband processor, and the like.

The antennas 1 and 2 are used for transmitting and receiving electromagnetic wave signals. Each antenna in the terminal device 100 may be used to cover a single or multiple communication bands. Different antennas may also be multiplexed to improve the utilization of the antennas. For example: the antenna 1 may be multiplexed into a diversity antenna of a wireless local area network. In other embodiments, the antenna may be used in conjunction with a tuning switch.

The mobile communication module 102A may provide a solution including 2G/3G/4G/5G wireless communication applied on the terminal device 100. The mobile communication module 102A may include at least one filter, switch, power amplifier, low noise amplifier (low noise amplifier, LNA), etc. The mobile communication module 102A may receive electromagnetic waves from the antenna 1, perform processing such as filtering and amplifying the received electromagnetic waves, and transmit the processed electromagnetic waves to a modem processor for demodulation. The mobile communication module 102A may amplify the signal modulated by the modem processor, and convert the signal into electromagnetic waves through the antenna 1 to radiate. In some embodiments, at least some of the functional modules of the mobile communication module 102A may be disposed in the processor 101. In some embodiments, at least some of the functional modules of the mobile communication module 102A may be provided in the same device as at least some of the modules of the processor 101.

The modem processor may include a modulator and a demodulator. The modulator is used for modulating the low-frequency baseband signal to be transmitted into a medium-high frequency signal. The demodulator is used for demodulating the received electromagnetic wave signal into a low-frequency baseband signal. The demodulator then transmits the demodulated low frequency baseband signal to the baseband processor for processing. The low frequency baseband signal is processed by the baseband processor and then transferred to the application processor. The application processor outputs sound signals through audio devices (not limited to speakers 106A, receivers 106B, etc.) or displays images or video streams through the display 103. In some embodiments, the modem processor may be a stand-alone device. In other embodiments, the modem processor may be provided in the same device as the mobile communication module 102A or other functional module, independent of the processor 101.

The wireless communication module 102B may provide solutions for wireless communication including wireless local area network (wireless local area networks, WLAN) (e.g., wireless fidelity (wireless fidelity, wi-Fi) network), bluetooth (BT), global navigation satellite system (global navigation satellite system, GNSS), frequency modulation (frequency modulation, FM), near field wireless communication technology (near field communication, NFC), infrared technology (IR), etc., applied to the terminal device 100. The wireless communication module 102B may be one or more devices that integrate at least one communication processing module. The wireless communication module 102B receives electromagnetic waves via the antenna 2, modulates the electromagnetic wave signals, filters the electromagnetic wave signals, and transmits the processed signals to the processor 101. The wireless communication module 102B may also receive a signal to be transmitted from the processor 101, frequency modulate it, amplify it, and convert it into electromagnetic waves via the antenna 2.

In some embodiments, antenna 1 and mobile communication module 102A of terminal device 100 are coupled, and antenna 2 and wireless communication module 102B are coupled, such that terminal device 100 may communicate with a network and other devices via wireless communication techniques. The wireless communication techniques may include the Global System for Mobile communications (global system for mobile communications, GSM), general packet radio service (general packet radio service, GPRS), code division multiple access (code division multiple access, CDMA), wideband code division multiple access (wideband code division multiple access, WCDMA), time division code division multiple access (time-division code division multiple access, TD-SCDMA), long term evolution (long term evolution, LTE), BT, GNSS, WLAN, NFC, FM, and/or IR techniques, among others. The GNSS may include a global satellite positioning system (global positioning system, GPS), a global navigation satellite system (global navigation satellite system, GLONASS), a beidou satellite navigation system (beidou navigation satellite system, BDS), a quasi zenith satellite system (quasi-zenith satellite system, QZSS) and/or a satellite based augmentation system (satellite based augmentation systems, SBAS).

As shown in fig. 1, in some implementations, the terminal device 100 may further include an external memory interface 110, an internal memory 104, a universal serial bus (universal serial bus, USB) interface, a charge management module 112, a power management module 113, a battery 114, an audio module 106, a speaker 106A, a receiver 106B, a microphone 106C, an earphone interface 106D, a sensor module 105, keys 109, a motor, an indicator 108, a camera 107, a subscriber identity module (subscriber identification module, SIM) card interface, and the like.

The charge management module 112 is configured to receive a charge input from a charger. The charger can be a wireless charger or a wired charger. In some wired charging embodiments, the charging management module 112 may receive a charging input of a wired charger through the USB interface 111. In some wireless charging embodiments, the charging management module 112 may receive wireless charging input through a wireless charging coil of the terminal device 100. The charging management module 112 may also supply power to the terminal device 100 through the power management module 113 while charging the battery 114.

The power management module 113 is used for connecting the battery 114, and the charge management module 112 and the processor 101. The power management module 113 receives input from the battery 114 and/or the charge management module 112 and provides power to the processor 101, the internal memory 104, the external memory, the display 103, the camera 107, the wireless communication module 102B, and the like. The power management module 113 may also be configured to monitor the capacity of the battery 114, the number of cycles of the battery 114, and the state of health (leakage, impedance) of the battery 114. In other embodiments, the power management module 113 may also be provided in the processor 101. In other embodiments, the power management module 113 and the charge management module 112 may be disposed in the same device.

The external memory interface 110 may be used to connect an external memory card, such as a Micro SD card, to realize expansion of the memory capability of the terminal device 100. The external memory card communicates with the processor 101 through an external memory interface 110 to implement data storage functions. For example, files such as music, video streams, etc. are stored in an external memory card.

The internal memory 104 may be used to store computer-executable program code that includes instructions. The processor 101 executes various functional applications of the terminal device 100 and data processing by executing instructions stored in the internal memory 104.

The internal memory 104 may also store one or more computer programs corresponding to the method for determining a user attribute of a terminal device provided in the embodiments of the present application.

The terminal device 100 may implement audio functions through the audio module 106, the speaker 106A, the receiver 106B, the microphone 106C, the earphone interface 106D, the application processor 101, and the like. Such as music playing, recording, etc.

The keys 109 include a power on key, a volume key, and the like. The keys 109 may be mechanical keys 109. Or may be a touch key 109. The terminal device 100 may receive key 109 inputs, generating key signal inputs related to user settings and function control of the terminal device 100.

The indicator 108 may be an indicator light, which may be used to indicate a state of charge, a change in charge, a message indicating a missed call, a notification, etc.

The SIM card interface is used for connecting the SIM card. The SIM card may be inserted into or withdrawn from the SIM card interface to enable contact and separation with the terminal apparatus 100. The terminal device 100 may support 1 or N SIM card interfaces, N being a positive integer greater than 1. The SIM card interface may support Nano SIM cards, micro SIM cards, etc. The same SIM card interface can be used to insert multiple cards simultaneously. The types of the plurality of cards may be the same or different. The SIM card interface may also be compatible with different types of SIM cards. The SIM card interface may also be compatible with external memory cards. The terminal device 100 interacts with the network through the SIM card to realize functions such as call and data communication. In some embodiments, the terminal device 100 employs esims, namely: an embedded SIM card. The eSIM card can be embedded in the terminal device 100 and cannot be separated from the terminal device 100.

The sensor module 105 in the terminal device 100 may include a touch sensor, a pressure sensor, a gyro sensor, a barometric sensor, a magnetic sensor, an acceleration sensor, a distance sensor, a proximity sensor, an ambient light sensor, a fingerprint sensor, a temperature sensor, a bone conduction sensor, etc. to implement sensing and/or acquisition functions for different signals.

It is to be understood that the configuration illustrated in the present embodiment does not constitute a specific limitation on the terminal device 100. In other embodiments, terminal device 100 may include more or fewer components than shown, or certain components may be combined, or certain components may be split, or different arrangements of components. The illustrated components may be implemented in hardware, software, or a combination of software and hardware.

The hardware structure of the terminal device provided in the embodiment of the present application is described above through fig. 1. The method for determining the user attribute of the terminal equipment provided by the embodiment of the application is specifically described below.

The method for determining the user attribute of the terminal device is used for determining the user attribute of the first terminal device according to the first data of the first terminal device. The first data includes basic parameters of the corresponding terminal device and usage data generated after the corresponding terminal device is activated and used. The user attributes include at least one of: user gender, user age group.

Referring to fig. 2, a flowchart of a method for determining a user attribute of a terminal device according to an embodiment of the present application is provided. As shown in fig. 2, the method includes S201-S203.

S201, acquiring first data of a first terminal device.

As described above, the basic parameters of the terminal device include at least one of the following: color, model, system version. The usage data of the terminal device at least comprises one of the following: application installation information, application use information, use time period, camera use frequency, camera setting parameters and system setting parameters. And will not be described in detail herein.

It should be noted that, in the embodiment of the present application, the acquired data is all performed under the premise of being agreed by the user, and the problem will not be described in detail later.

S202, inputting first data of the first terminal equipment into the attribute determination model. The attribute determination model is generated by training a preset model by using sample data in advance. The sample data includes first data of a plurality of terminal devices and user attributes. The preset model is a classification model.

S203, determining the user attribute of the first terminal equipment according to the output result of the attribute determination model.

In other words, the embodiment of the application trains the preset model in advance by using the sample data to obtain the attribute determination model. And inputting the first data of the first terminal equipment into the attribute determination model to obtain the user attribute of the first terminal equipment.

In some embodiments, the preset model may be a linear classification model, i.e. a classification model in which the decision boundary is a linear boundary. In other embodiments, the pre-set model may also be a tree model classifier, such as GBDT (Gradient Boosting Decison Tree, gradient-lifting decision tree).

In embodiments of the present application, the process of training the preset model using the sample data may include super-parameter selection and offline evaluation. The following is a detailed description.

Referring to fig. 3, a flowchart of a model training method according to an embodiment of the present application is provided. As shown in FIG. 3, the model training method may include S301-S306.

S301, dividing sample data into a training set and a testing set.

Wherein the training set may comprise first data and user attributes of a part of the terminal devices and the test set may comprise first data and user attributes of another part of the terminal devices. The method for dividing the training set and the test set is not limited in the embodiment of the application.

S302, selecting a new group of super parameters to train a preset model on the training set.

Hyper-parameters refer to parameters that are manually set before starting training a pre-set model. It should be noted that, in the model training method provided in the embodiment of the present application, the hyper-parameter combinations are continuously optimized during the training process, so as to obtain the hyper-parameter combinations with the best performance on the training set.

S303, inputting the test set into the trained preset model to obtain a test result.

S304, evaluating the test result in an offline environment to obtain an evaluation result.

The evaluation index may be accuracy, average absolute error, etc., and is not limited herein.

S305, judging whether the iteration times meet the preset times. If yes, executing S306; if not, S301 is performed.

In this example, performing S301-S304 once may be referred to as iterating once. That is, the number of iterations may refer to the number of times S301-S304 are performed. The preset number of times and the preset condition may be input in advance. Illustratively, the preset number of times may be one hundred, three hundred, etc., and the preset condition may be an accuracy of more than seventy percent, etc., which is not limited in this application.

In the embodiment of the present application, the determination condition may be other conditions, and this is merely illustrative, and does not represent the limitation of the present application.

S306, applying a group of super parameters corresponding to the optimal evaluation result to a preset model to obtain an attribute determination model.

The above is a process of training a preset model by using sample data provided in the embodiments of the present application. In the case where model training includes super-parameter selection and offline evaluation in the subsequent embodiments, the process may refer to S301 to S306 described above.

Based on the above description, it can be seen that the method for determining the user attribute of the terminal device according to the embodiment of the present application uses the machine learning model trained by the sample data to realize that the user attribute of the terminal device can still be determined according to the basic parameters and the usage data of the terminal device without including the user attribute in the terminal device, and has higher accuracy.

In order to reduce the pressure of data processing and the efficiency of data processing, the method for determining the user attribute of the terminal equipment provided by the embodiment of the invention can train the feature data after the first data processing in the model training stage. In the model use stage, the first data of the first terminal equipment can be subjected to feature extraction to obtain the feature data of the first terminal equipment, and then the feature data of the first terminal equipment is input into the model to obtain the user attribute of the first terminal equipment. In this way, the accuracy and efficiency of determining the user attributes may be improved.

The feature extraction may include feature derivation and clustering, which are described below.

In order to facilitate understanding, in the following description, a user attribute to be determined by the method for determining a user attribute of a terminal device provided in the embodiment of the present application is a user gender. The terminal equipment is a mobile phone. The base parameter in the first data of the first terminal device comprises a time period during which the terminal device is used daily.

Feature derivation will be described first. Feature derivation may also be referred to as feature construction, meaning construction of new features from raw data. In the embodiment of the present application, the process of extracting the features of the basic parameters of the first data of the first terminal device or extracting the features of the basic parameters of each first data in the sample data may refer to feature derivation. In order to avoid redundancy, the basic parameters of the first data of the first terminal device will be referred to as the basic parameters of the first terminal device in the following description.

For example, the base parameters of the first terminal device include a time period during which the terminal device is used daily. The basic parameters are not very relevant to whether the user of the terminal device is male or female. In other words, the correlation between the time period in which the terminal device is used every day and the sex of the user of the terminal device is small. And the earliest and latest times of day that the terminal device is used have a certain correlation with the sex of the user of the terminal device. Thus, the "period of time in which the terminal device is used every day" can be feature-derived, resulting in "earliest time and latest time in which the terminal device is used every day". Of course, this is merely an example and is not intended to limit the present application.

Clustering is described below. Clustering refers to the process of dividing a collection of physical or abstract objects into multiple classes made up of similar objects. In this embodiment of the present application, feature extraction of usage data in the first data of the first terminal device or feature extraction of usage data of each first data in the sample data may refer to clustering. In order to avoid redundancy, the usage data in the first data of the first terminal device will be referred to as usage data of the first terminal device in the following description.

The usage data of the first terminal device includes application installation information. The application installation information may include the number of applications installed in the terminal device, application names, and the like. It will be appreciated that the number of applications in the application market is very large, and the data throughput may be large if the application installation information is directly used to train the model or to input a trained model. Therefore, the method for determining the user attribute of the terminal equipment provided by the embodiment of the application can classify the applications with the quantity greater than the first threshold value in the application market in advance, so as to obtain a plurality of application categories. And determining application categories included in the first terminal equipment according to the application installation information, the application installation quantity under each application category, the application use times, the application use duration and the like. Therefore, the application installation information can be clustered into information of different application categories, and the data processing amount and the processing complexity are greatly reduced.

In the embodiment of the application, when the model is trained, different training strategies can be selected according to the activation time length of the terminal equipment. For example, a terminal device with a longer activation period has a larger data size of the first data or the feature data, so that a strategy with smaller complexity and less calculation force requirement can be adopted for model training. And the terminal equipment with shorter activation duration has smaller data size of the first data or the characteristic data, so that the model training can be performed by adopting a strategy with larger complexity and calculation force requirements.

In the embodiment of the present application, the terminal device with a longer activation duration may refer to a terminal device with an activation time greater than 14 days. The terminal device with a shorter activation time may refer to a terminal device with an activation time less than 14 days.

The training strategy adopted for the first data of the terminal equipment with longer activation duration when the model is trained is exemplified below. It should be understood that this process of training the model is performed before S201 described above.

Referring to fig. 4, a flowchart of another model training method according to an embodiment of the present application is provided. As shown in fig. 4, the method includes S401 to S403.

S401, acquiring sample data.

Wherein the sample data comprises first data of a plurality of terminal devices and user attributes. As described above, the embodiment of the present application may perform feature extraction on the first data of each terminal device to obtain feature data of each terminal device, so as to reduce data processing amount and improve data processing efficiency.

Thus, in one possible design, the sample data may also include characteristic data of a plurality of terminal devices as well as user attributes. The feature data is data generated after feature extraction of the first data of the terminal equipment.

In addition, in the embodiment of the present application, in the sample data, the first data of the terminal device with a longer duration is activated, and the time scale may be 60 days to 90 days before the acquisition time point. The first data of the terminal device with a shorter duration may be activated, and the time size may be 7 days to 14 days before the acquisition time point. Thus, the reliability of the acquired first data is improved.

S402, acquiring a first sample in sample data. The activation time length of each terminal device corresponding to the first sample is longer than the first time length.

That is, the first samples may be obtained from the sample data by determining whether the activation duration of the terminal device corresponding to each first data in the sample data is longer than the first duration.

It can be understood that the first sample includes feature data of the terminal device with a longer activation time and user attributes.

S403, training a first model in the preset models by using the first samples, and generating a second model in the attribute determination model. The first model is a tree model classifier or a linear classification model.

That is, the preset model in the embodiment of the present application may include a plurality of models, where the first model is one of the models. The attribute determination model may also include a plurality of models, one of which is the second model. The process of training the first model using the first sample may refer to S301-S306 described above, and will not be described here.

The characteristic data of the part of terminal equipment has the characteristics of large data volume and stable data distribution because the first sample corresponds to the terminal equipment with longer activation time. Therefore, the first model can be trained by using the hive+MMLSpark architecture, so that the calculation force requirement, the efficiency, the resource consumption and the time consumption of model training are considered.

The processing procedure of the first data of the first terminal device may be as shown in fig. 5 on the basis of the above-mentioned S401-S403. Referring to fig. 5, a flowchart of a method for determining a user attribute of another terminal device according to an embodiment of the present application is provided. As shown in fig. 5, the method includes S501-S503.

S501, extracting features of the first data of the first terminal equipment to obtain the feature data of the first terminal equipment.

The description of feature extraction can be found in the foregoing embodiments, and will not be described herein.

S502, when the activation time of the first terminal equipment is longer than the first time, inputting the characteristic data of the first terminal equipment into the second model.

That is, when the activation time period of the first terminal device is longer than the first time period, the feature data of the first terminal device is processed by using the second model trained by the feature data of the terminal device whose activation time period is longer than the first time period. In this way, it is advantageous to increase the accuracy of the determined user attributes.

S503, determining the user attribute of the first terminal equipment according to the output result of the second model.

Thus, the user attribute of the first terminal device can be obtained.

The above is an exemplary description of the training strategy adopted for activating the first data of the terminal device with longer duration. The training strategy adopted for the first data of the terminal equipment with longer activation duration is further exemplified below. It should be understood that this process of training the model is also performed before S201 described above.

Referring to fig. 6, a flowchart of another model training method according to an embodiment of the present application is provided. As shown in fig. 6, the method includes S601-S606.

S601, acquiring sample data.

The description of the sample data may be referred to S401 above, and will not be described herein.

S602, acquiring a second sample in the sample data. And the activation time length of each terminal device corresponding to the second sample is smaller than or equal to the first time length.

That is, the second sample may be obtained from the sample data by determining whether the activation duration of the terminal device corresponding to each first data in the sample data is less than or equal to the first duration.

It can be understood that the second sample includes feature data of the terminal device with a shorter activation time and the user attribute.

S603, inputting the second sample into the tree model classifier for multiple iterations to obtain leaf node data of each weak classifier in the tree model classifier. The leaf node data is a feature combination of the feature data in the second sample.

The tree model classifier is a classification model which needs to determine the category to which data belongs through multi-level discrimination. In one possible design, the tree model classifier is a GBDT, i.e., a gradient-lifting decision tree.

Please refer to fig. 7, which is a schematic diagram of a tree model classifier according to an embodiment of the present application. As shown in fig. 7, the tree model classifier includes m trees, each tree including n leaves. Wherein each dashed box represents the structure of 1 tree. The different circles, except for the circle of the last row, represent different feature discrimination conditions. The last row, the circle of Leaf1_1 through leaf_n, represents Leaf node data.

After the second sample is input into the data inlet of the tree model classifier, the tree model classifier firstly classifies each feature data in the second sample onto corresponding trees, and then combines each feature in the feature data step by step according to the feature discrimination conditions in the corresponding trees, so as to finally obtain leaf node data.

Taking the use time period of the terminal device included in the feature data and application use information as an example, the root node of the tree1 in fig. 7, that is, the discrimination condition 1 may be whether to use the trip navigation application, if yes, the trip navigation application falls on the discrimination condition 2. If not, then it falls on criterion 3. The criterion 2 may be whether a music application is used, if so, falling on leaf1_1, and if not, falling on leaf1_2. The judging condition 3 can be whether a camera application is used or not, if yes, the judging condition falls on the judging condition 4, and if not, the classifying result is not output without processing. The discrimination condition 4 may be a beauty grade set when the camera application is used, if the beauty grade is higher, the beauty grade falls in Leaf1_3, and if the beauty grade is not higher, the classification result is not output without processing.

If the feature data of a certain terminal device is input into the tree model classifier and then finally output into the Leaf1_3, the fact that the user does not use the travel navigation application, the camera application is used, and the beauty grade set when the camera application is used is higher is indicated.

It can be seen that the leaf node data obtained after the feature data is input into the tree model classifier is a combination of various different features in the feature data. That is, a tree model classifier may be used to combine features from different features in the feature data. Thus, the model training efficiency is improved.

It should be understood that, in the embodiment of the present application, if the second sample includes p pieces of training data, training a tree model classifier with m trees and n leaves of each tree may obtain which leaf node data of each tree each piece of training data falls on, so that p×m pieces of leaf node data may be obtained.

Here, the leaf node data is only used to indicate what number of leaf data of what tree, and does not include actual features, so the leaf node data needs to be further encoded to be used to train the preset model.

Illustratively, the resulting leaf node data may form a matrix of p×m as follows:

S604, performing one-bit effective coding on each leaf node data to obtain coded data.

Wherein a one-bit valid code, which may also be referred to as a one-hot code, refers to the encoding of N states with N-bit state registers, each state having its own independent register bit, and at any time only one of the bits is valid.

Along with the above example, the output Leaf node data is leaf1_3, that is, the user of the terminal device does not use the travel navigation application, uses the camera application, and sets a higher beauty level when using the camera application.

The user of the terminal equipment uses the travel navigation application to take a value of yes or no. Whether the camera application is used or not has a value of yes or no, and the beauty grade set when the camera application is used has a low value, a higher value and a high value. The following illustrates a one-hot encoding process.

For whether or not the travel navigation class application is used during use, represented by 10 is, and 01 is, represented by no. For whether a camera application is used, represented by 10 is, and 01 is, represented by no. For the beauty grade set when using the camera application, 100 represents low, 010 represents high, and 001 represents high. The Leaf node data Leaf1_3 may be 0110010 after being one-hot encoded. The order from the highest to the lowest, 01 represents an unused travel navigation class application, 10 represents a used camera application, 010 represents a set beauty level higher.

Of course, other values may be used for the one-hot encoding, and the above is merely exemplary and not intended to limit the present application.

The one-hot encoding process may also be, for example, that the feature data falls after the leaf node data of a certain tree after being input into the tree model classifier, such as f of the first tree ₀₀ Leaf node data, an n-dimensional vector representing n leaves of the tree may be created. F in the vector ₀₀ Dimension is set to 1, divided by f ₀₀ All other dimensions are set to 0. Thus, if the second sample includes p pieces of training data, training the tree model classifier with m trees and n leaves of each tree, and then obtaining leaf node data corresponding to each piece of training data, where the feature dimension is m×n.

And after the data of each leaf node are coded, the coded data can be obtained. It will be appreciated that the encoded data will be much smaller in size than the data after the full permutation of the characteristic data of the second sample. Thus, the data processing efficiency is improved.

And S605, combining the coded data with each characteristic data in the second sample to obtain a third sample.

Therefore, the dimension of the sample data can be further improved, and the training efficiency of the model is improved.

S606, training a third model in the preset models by using the third sample, and generating a fourth model in the attribute determination model. The fourth model is a linear classification model.

That is, the preset model in the embodiment of the present application may include a plurality of models, and the third model is one of the models. The attribute determination model may also include a plurality of models, one of which is a fourth model. The process of training the third model using the third sample may refer to S301 to S306 described above, and will not be described here.

It should be noted that, because the second sample corresponds to the terminal device with shorter activation duration, the characteristic data of this part of terminal device has the characteristics of small data size and larger change of data distribution along with the use time. Therefore, the hive partitioning transformation data structure+python architecture can be adopted to perform feature crossing through the tree model classifier, and the new feature (namely leaf node data) after crossing is subjected to one-hot coding and then is converted into the new feature (namely coded data). After combining the encoded data and the original features (i.e., the feature data in the second sample), a new training feature (i.e., the third sample) is constructed, and the third model is trained. Thus, the model training efficiency and the model accuracy can be improved.

The above is an exemplary description of the training strategy adopted by the first data of the terminal device with shorter activation duration.

The processing procedure of the first data of the first terminal device may be as shown in fig. 8 on the basis of the above S601-S606. Referring to fig. 8, a flowchart of a method for determining a user attribute of another terminal device according to an embodiment of the present application is provided. As shown in fig. 8, the method includes S801 to S807.

S801, extracting features of first data of first terminal equipment to obtain feature data of the first terminal equipment.

S802, when the activation time length of the first terminal equipment is smaller than or equal to the first time length, inputting the feature data of the first terminal equipment into a tree model classifier to obtain a feature classification result.

The working principle of the tree model classifier can be seen in the foregoing embodiments, and will not be described herein. It should be noted that, the feature classification result is the leaf node data.

S803, performing one-bit effective coding on the feature classification result to obtain a feature coding result.

The description of the one-bit valid code can be found in the previous embodiments, and will not be repeated here.

S804, combining the feature coding result with the feature data of the first terminal equipment to obtain a feature combining result.

S805, taking the sequence of the coded data and the sequence of the feature data in the third sample as references, and performing feature alignment on the feature merging result to obtain a feature alignment result.

The feature data of the first terminal device may include features not included in the sample data due to the limited data size of the sample data. Therefore, before inputting the feature merging result into the fourth model, the feature alignment may be performed on the feature merging result with reference to the sequence of the encoded data in the third sample and the sequence of the feature data.

In the feature merging result, features which do not appear in the encoded data and the feature data of the second sample may be deleted, and the missing features may be supplemented by 0 compared with the encoded data and the feature data of the second sample, thereby completing the feature alignment process.

S806, inputting the characteristic alignment result into the fourth model.

In the embodiment of the application, the feature alignment result is usually a high-dimensional sparse feature, so that a regular term can be added to fit the feature when the fourth model is input. The regular terms may be weights corresponding to the high-dimensional sparse features generated when the second sample is used to train the third model, which are not specifically limited herein.

S807, determining the user attribute of the first terminal device according to the output result of the fourth model.

Thus, the user attribute of the first terminal device can be obtained.

In some other embodiments, the training of the preset model using the sample data and the determining of the user attribute of the first terminal device using the attribute determining model may be performed simultaneously. Illustratively, after receiving the full data packet, sample data and data of the user attribute to be determined may be separated from the full data packet; and processing the data of the user attribute to be determined by using a model trained by the sample data, thereby improving the efficiency of data processing.

According to the method for determining the user attribute of the terminal equipment, the machine learning model trained by the sample data is used, so that the user attribute of the terminal equipment can still be determined according to the basic parameters and the use data of the terminal equipment under the condition that the terminal equipment does not comprise the user attribute, and the accuracy is high.

Fig. 9 is a schematic diagram of an electronic device 900 according to an embodiment of the present application. The electronic device 900 may be any of the above examples, for example, the electronic device 900 may be a mobile phone, a computer, or the like. For example, as shown in fig. 9, the electronic device 900 may include: a processor 901 and a memory 902. The memory 902 is used to store computer-executable instructions. For example, in some embodiments, the processor 901 may be configured to cause the electronic device 900 to perform any of the functions of the electronic device in the above embodiments when executing the instructions stored in the memory 902, so as to implement the user attribute determining method of any of the terminal devices in the above examples.

It should be noted that, all relevant contents of each step related to the above method embodiment may be cited to the functional description of the corresponding functional module, which is not described herein.

Fig. 10 shows a schematic diagram of the components of a chip system 1000. The chip system 1000 may be provided in an electronic device. For example, the system on chip 1000 may be provided in a mobile phone. By way of example, the chip system 1000 may include: a processor 1001 and a communication interface 1002 for supporting the electronic device to implement the functions referred to in the above embodiments. In one possible design, the chip system 1000 may further include a memory to hold program instructions and data necessary for the electronic device. The chip system can be composed of chips, and can also comprise chips and other discrete devices. It should be noted that, in some implementations of the present application, the communication interface 1002 may also be referred to as an interface circuit.

The present application further provides a computer storage medium having stored therein computer instructions which, when executed on a terminal device, cause the terminal device to perform the above-mentioned related method steps to implement the method in the above-mentioned embodiments.

The present application also provides a computer program product which, when run on a computer, causes the computer to perform the above-mentioned related steps to implement the method in the above-mentioned embodiments.

In addition, embodiments of the present application also provide an apparatus, which may be specifically a chip, a component, or a module, and may include a processor and a memory connected to each other; the memory is configured to store computer-executable instructions, and when the device is operated, the processor may execute the computer-executable instructions stored in the memory, so that the chip performs the methods in the above method embodiments.

The terminal device, the computer storage medium, the computer program product, or the chip provided in the embodiments of the present application are used to execute the corresponding methods provided above, so that the beneficial effects that can be achieved by the terminal device, the computer storage medium, the computer program product, or the chip can refer to the beneficial effects in the corresponding methods provided above, and are not described herein.

The above description has been made mainly from the point of view of the electronic device. To achieve the above functions, it includes corresponding hardware structures and/or software modules that perform the respective functions. Those of skill in the art will readily appreciate that the elements and algorithm steps of the examples described in connection with the embodiments disclosed herein may be implemented as hardware or combinations of hardware and computer software. Whether a function is implemented as hardware or computer software driven hardware depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application.

The embodiments of the present application may divide functional modules of devices involved therein according to the above method examples, for example, each functional module may be divided corresponding to each function, or two or more functions may be integrated in one processing module. The integrated modules may be implemented in hardware or in software functional modules. It should be noted that, in the embodiment of the present application, the division of the modules is schematic, which is merely a logic function division, and other division manners may be implemented in actual implementation.

The functions or acts or operations or steps and the like in the embodiments described above may be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using a software program, it may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the processes or functions described in accordance with embodiments of the present application are produced in whole or in part. The computer may be a general purpose computer, a special purpose computer, a computer network, or other programmable apparatus. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium, for example, the computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by a wired (e.g., coaxial cable, fiber optic, digital subscriber line (digital subscriber line, DSL)) or wireless (e.g., infrared, wireless, microwave, etc.). The computer readable storage medium may be any available medium that can be accessed by a computer or a data storage device including one or more servers, data centers, etc. that can be integrated with the medium. The usable medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a magnetic tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a Solid State Disk (SSD)), or the like.

Although the present application has been described in connection with specific features and embodiments thereof, it will be apparent that various modifications and combinations can be made without departing from the spirit and scope of the application. Accordingly, the specification and drawings are merely exemplary illustrations of the present application as defined in the appended claims and are considered to cover any and all modifications, variations, combinations, or equivalents that fall within the scope of the present application. It will be apparent to those skilled in the art that various modifications and variations can be made in the present application without departing from the spirit or scope of the application. Thus, if such modifications and variations of the present application fall within the scope of the claims and the equivalents thereof, the present application is intended to include such modifications and variations as well.

Claims

1. A method for determining a user attribute of a terminal device, comprising:

acquiring first data of first terminal equipment; the first data comprise basic parameters of the first terminal equipment and use data generated after the first terminal equipment is activated and used; the basic parameters of the first terminal equipment comprise the activation duration of the first terminal equipment;

respectively carrying out feature derivation on basic parameters in first data of the first terminal equipment, and clustering usage data in the first data of the first terminal equipment to obtain feature data of the first terminal equipment;

When the activation time of the first terminal equipment is longer than the first time, inputting the characteristic data of the first terminal equipment into a second model; determining the user attribute of the first terminal equipment according to the output result of the second model; the second model is generated by training a first model in advance by using a first sample, the activation time length of terminal equipment corresponding to the first sample is longer than the first time length, and the first model is a tree model classifier or a linear classification model; the user attributes include at least one of: user gender, user age group;

when the activation time length of the first terminal equipment is smaller than the first time length, inputting the characteristic data of the first terminal equipment into a tree model classifier to obtain a characteristic classification result; carrying out one-bit effective coding on the feature classification result to obtain a feature coding result; combining the feature coding result with the feature data of the first terminal equipment to obtain a feature combining result; taking the sequence of the coded data and the sequence of the feature data in the third sample as references, and carrying out feature alignment on the feature merging result to obtain a feature alignment result; inputting the feature alignment result into a fourth model; determining the user attribute of the first terminal equipment according to the output result of the fourth model;

The third sample is generated by performing first training by using the second sample in advance; the activation time length of the terminal equipment corresponding to the second sample is smaller than the first time length; the first training comprises the steps of inputting the second sample into a tree model classifier for multiple iterations to obtain leaf node data of each weak classifier in the tree model classifier; the leaf node data are feature combinations of the feature data in the second sample; carrying out one-bit effective coding on each leaf node data to obtain coded data; combining the encoded data with the feature data in the second sample;

the fourth model is generated by training a third model in advance by using the third sample; the fourth model is a linear classification model.

2. The method of claim 1, wherein the tree model classifier is a gradient-lifting decision tree.

3. The method according to claim 1, wherein said characterizing the base parameters in the first data of the first terminal device when the base parameters in the first data of the first terminal device comprise a period of time during which the terminal device is used each day, comprises:

And determining the earliest time and the latest time of the daily use of the terminal equipment according to the daily use time period of the terminal equipment.

4. The method according to claim 1, wherein when the usage data in the first data of the first terminal device includes application installation information, the clustering the usage data in the first data of the first terminal device includes:

classifying applications with the number greater than a first threshold value in the application market to obtain a plurality of application categories;

and determining the application category included in the first terminal equipment according to the application installation information.

5. The method according to claim 1, characterized in that the basic parameters of the terminal device comprise at least one of the following: color, model, system version.

6. The method of claim 1, wherein the usage data of the terminal device comprises at least one of: application installation information, application use information, use time period, camera use frequency, camera setting parameters and system setting parameters.

7. An electronic device comprising one or more processors and one or more memories; the one or more memories coupled to the one or more processors, the one or more memories storing computer instructions;

The computer instructions, when executed by the one or more processors, cause the electronic device to perform the user attribute determination method of a terminal device of any of claims 1-6.

8. A computer readable storage medium, characterized in that the computer readable storage medium comprises computer instructions which, when run, perform the method of determining a user attribute of a terminal device according to any of claims 1-6.

9. A computer program product comprising instructions which, when run on a computer, cause the computer to perform the method of determining a user attribute of a terminal device according to any of claims 1-6.