CN107807997A - User's portrait building method, device and computing device based on big data - Google Patents

Abstract

The invention discloses a kind of user's portrait building method, device, computing device and computer-readable storage medium based on big data.Wherein method includes：Obtain user's geographical location information and user operates the behavioural information of application program；Position point sequence is generated according to user's geographical location information；Position point sequence is learnt using the first deep learning algorithm, obtains user's geographic location feature vector corresponding to user；Behavioural information is learnt using the second deep learning algorithm, obtains user behavior characteristic vector；Drawn a portrait based on user's geographic location feature vector sum user behavior characteristic vector structuring user's.Technical scheme provided by the invention, user's portrait based on big data construction can fully reflect the feature of user, so as to easily provide the user the service of more becoming more meticulous according to user's portrait, overcome the metadata for relying only on user in the prior art, for example, the low-quality defect of user's portrait caused by the structuring user's such as age, sex, income portrait.

Description

User portrait construction method, device and computing equipment based on big data

technical field

The present invention relates to the technical field of data processing, in particular to a user portrait construction method, device, computing device and computer storage medium based on big data.

Background technique

With the continuous development of network and information technology, various platform applications are born. In order to improve the various functions of the platform, to facilitate the platform to provide users with better services, and to protect the interests of the platform, it is necessary to understand and analyze various aspects of users. information. At present, user information is often learned by constructing user portraits.

Existing user portrait construction methods mainly construct user portraits based on the user’s unary data, such as gender, age, occupation, constellation, height, weight, shopping type, brand preference and/or income, etc., and do not make full use of the factors The big data generated by the Internet, although the user portrait constructed by using the unary data can guide the platform, but it cannot fully reflect the characteristics of the user, so that the platform cannot provide users with refined services, and it is impossible for the user to be fraudulent, etc. Behavior to protect the interests of the platform.

Contents of the invention

In view of the above problems, the present invention is proposed to provide a big data-based user portrait construction method, device, computing device, and computer storage medium that overcome the above problems or at least partially solve the above problems.

According to one aspect of the present invention, a method for constructing user portraits based on big data is provided, including:

Obtain the user's geographical location information and the behavior information of the user's operation of the application;

Generate a location point sequence according to the user's geographic location information;

Using the first deep learning algorithm to learn the location point sequence to obtain the user's geographic location feature vector corresponding to the user;

Using the second deep learning algorithm to learn the behavior information to obtain the user behavior feature vector;

Construct user portraits based on user geographic location feature vectors and user behavior feature vectors.

Optionally, generating the location point sequence according to the user's geographic location information further includes:

Combining geographic location information to generate location information tracks;

Obtain information around the location point based on the location point on the location information track, and generate a sequence of location points;

Using the first deep learning algorithm to learn the location point sequence, and obtaining the user's corresponding user geographic location feature vector further includes:

The trained time recursive neural network is used to perform sequence learning on the location point sequence, and the user's geographic location feature vector corresponding to the user is obtained.

Optionally, using the second deep learning algorithm to learn the behavior information, and obtaining the user behavior feature vector further includes:

According to the behavior information, the behavior track of the user operating the application is obtained;

Generate a behavior sequence according to the behavior trajectory of the user operating the application;

The behavior sequence is learned sequentially by using the trained time recurrent neural network, and the user behavior feature vector is obtained.

Optionally, the method further includes: generating a location consistency feature vector according to the user's geographic location information;

Constructing user portraits based on user geographic location feature vectors and user behavior feature vectors further includes:

Perform verification processing on the user geographic location feature vector according to the location consistency feature vector;

A user profile is constructed according to the verified user geographic location feature vector and user behavior feature vector.

Optionally, after constructing the user portrait, the method further includes: training a user interest and hobby model based on the user portrait sample;

Input the user portrait of the user to be analyzed into the user interest model for analysis, and obtain the user's interest and hobbies.

Optionally, after constructing the user portrait, the method further includes: analyzing whether the user has abnormal behavior based on the user portrait.

Optionally, the geographical location information of the user includes: IP information, WIFI information, base station information and/or GPS information.

Optionally, the behavior information includes: the user's operation on the application program and the time spent on the page of the application program.

According to another aspect of the present invention, a device for constructing user portraits based on big data is provided, including:

An acquisition module, adapted to acquire user geographic location information and behavior information of user operating applications;

The first generation module is adapted to generate a location point sequence according to the user's geographic location information;

The first learning module is adapted to use the first deep learning algorithm to learn the location point sequence, and obtain the user's geographic location feature vector corresponding to the user;

The second learning module is adapted to use a second deep learning algorithm to learn behavior information to obtain user behavior feature vectors;

The user portrait construction module is suitable for constructing a user portrait based on user geographic location feature vectors and user behavior feature vectors.

Optionally, the first generating module further includes:

A combination processing unit is suitable for combining and processing the geographic location information to generate a location information track;

The first generation unit is adapted to obtain information around the location point based on the location point on the location information track, and generate a sequence of location points;

The first learning module is further adapted to: use the trained time recursive neural network to perform sequence learning on the location point sequence to obtain the user's geographic location feature vector corresponding to the user.

Optionally, the second study module further includes:

A behavior track generating unit, adapted to obtain the behavior track of the user operating the application program according to the behavior information;

A behavior sequence generation unit adapted to generate a behavior sequence according to the behavior trajectory of the user operating the application program;

The second learning unit is adapted to use the trained time recurrent neural network to perform sequence learning on the behavior sequence to obtain the user behavior feature vector.

Optionally, the device further includes: a second generating module, adapted to generate a location consistency feature vector according to the user's geographic location information;

The user portrait construction module further includes:

The verification processing unit is suitable for performing verification processing on the user geographic location feature vector according to the location consistency feature vector;

The user portrait construction unit is adapted to construct a user portrait according to the verified user geographic location feature vector and user behavior feature vector.

Optionally, the device further includes: a training module, adapted to train a user interest model based on a user portrait sample;

The hobby analysis module is adapted to input the user profile of the user to be analyzed into the user hobby model for analysis, and obtain the user's hobbies.

Optionally, the device further includes: an abnormal behavior analysis module, adapted to analyze whether the user has abnormal behavior based on the user profile.

Optionally, the geographical location information of the user includes: IP information, WIFI information, base station information and/or GPS information.

Optionally, the behavior information includes: the user's operation on the application program and the time spent on the page of the application program.

According to yet another aspect of the present invention, a computing device is provided, including: a processor, a memory, a communication interface, and a communication bus, and the processor, the memory, and the communication interface complete mutual communication through the communication bus;

The memory is used to store at least one executable instruction, and the executable instruction causes the processor to execute the operation corresponding to the above-mentioned big data-based user portrait construction method.

According to another aspect of the present invention, a computer storage medium is provided, and at least one executable instruction is stored in the storage medium, and the executable instruction causes the processor to perform operations corresponding to the above-mentioned user portrait construction method based on big data.

According to the solution provided by the present invention, user geographic location information and user operation application behavior information are obtained, a sequence of location points is generated according to the geographic location information of the user, and the sequence of location points is learned by using the first deep learning algorithm to obtain the user's geographic location corresponding to the user. The location feature vector uses the second deep learning algorithm to learn the behavior information to obtain the user behavior feature vector, and constructs the user portrait based on the user geographic location feature vector and the user behavior feature vector. According to the technical solution provided by the present invention, the user portrait constructed based on big data can fully reflect the characteristics of the user, so that more refined services can be conveniently provided for the user according to the user portrait, and it overcomes the unary data that only relies on the user in the prior art, such as , The defects of low quality of user portraits caused by constructing user portraits such as age, gender, and income.

The above description is only an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention, it can be implemented according to the contents of the description, and in order to make the above and other purposes, features and advantages of the present invention more obvious and understandable , the specific embodiments of the present invention are enumerated below.

Description of drawings

Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiment. The drawings are only for the purpose of illustrating a preferred embodiment and are not to be considered as limiting the invention. Also throughout the drawings, the same reference numerals are used to designate the same parts. In the attached picture:

Fig. 1 shows a schematic flow chart of a method for constructing user portraits based on big data according to an embodiment of the present invention;

Fig. 2 shows a schematic flow chart of a method for constructing user portraits based on big data according to another embodiment of the present invention;

FIG. 3 shows a schematic structural diagram of a device for constructing user portraits based on big data according to an embodiment of the present invention;

Fig. 4 shows a structural schematic diagram of a user portrait construction device based on big data according to another embodiment of the present invention;

Fig. 5 shows a schematic structural diagram of a computing device according to an embodiment of the present invention.

Detailed ways

Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be embodied in various forms and should not be limited by the embodiments set forth herein. Rather, these embodiments are provided for more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

Fig. 1 shows a schematic flowchart of a method for constructing a user portrait based on big data according to an embodiment of the present invention. As shown in Figure 1, the method includes the following steps:

Step S100, acquiring the user's geographic location information and the behavior information of the user's operation of the application program.

Wherein, the user geographic location information may be reported by the user terminal. Optionally, the user geographic location information in this embodiment may include one or more of the following information: IP information, WIFI information, base station information, GPS information, Based on the above information, the current geographic location of the user can be determined. For example, the city where the user is located can be determined based on the IP address, the country where the user is located can be determined based on base station information, street information can be determined based on GPS information, etc. For example, a certain IP address can be obtained The address is 118.119.85.36. According to the obtained IP address query IP lookup table, it can be determined that the specific geographic location is Ziyang City, Sichuan Province. This is just an example and does not have any limiting effect. The geographic location of the user can be obtained according to the above four kinds of information, and no further examples are given here.

Behavior information of users operating applications refers to the behavior information generated when users operate applications. The behavior information here includes not only specific operations but also the time spent on the pages of applications. According to the behavior information of users operating applications, it can be Analyze whether the user's behavior is friendly, and analyze the information that the user cares about.

Step S101, generating a location point sequence according to the user's geographic location information.

After the user's geographic location information is obtained, a sequence of location points can be generated based on the obtained user's geographic location information. In this sequence of location points, the order of the location points is particularly important, and is sorted according to the order of time. For example Note that the location point sequence generated by obtaining the user's geographic location information in a day is: Beijing→Shanghai→Sanya→Beijing. Of course, this is just an example and does not have any limiting effect.

Step S102, using the first deep learning algorithm to learn the sequence of location points to obtain the feature vector of the user's geographic location corresponding to the user.

After the location point sequence is obtained according to step S101, the first deep learning algorithm can be used to learn the location point sequence. Here, the location point sequence is learned mainly to obtain the user's geographic location feature vector to represent the user's geographic location information in the form of a vector. , to facilitate subsequent construction of user portraits.

Step S103, using the second deep learning algorithm to learn the behavior information to obtain the user behavior feature vector.

After obtaining the behavior information of the user's operation application program according to step S100, the second deep learning algorithm can be used to learn the behavior information. Here, the learning of the behavior information is mainly to obtain the user behavior feature vector, which is used to represent the user operation application in the form of a vector. The behavior information of the program is convenient for subsequent construction of user portraits.

Step S104, constructing a user portrait based on the user's geographic location feature vector and the user behavior feature vector.

After obtaining the user geographic location feature vector and user behavior feature vector according to step S102 and step S103, a user portrait can be constructed according to the obtained user geographic location feature vector and user behavior feature vector, and the user portrait can fully reflect the characteristics of the user, It overcomes the defects in the existing technology that only rely on the user's unary data, such as age, gender, income, etc. to construct the user portrait, which leads to the low quality of the user portrait and cannot provide users with more refined services.

According to the method provided by the above-mentioned embodiments of the present invention, the user's geographic location information and the behavior information of the user's operation application program are obtained, a location point sequence is generated according to the user's geographic location information, and the location point sequence is learned by using the first deep learning algorithm to obtain the user's corresponding The user's geographic location feature vector is used to learn the behavior information using the second deep learning algorithm to obtain the user behavior feature vector, and the user portrait is constructed based on the user's geographic location feature vector and the user behavior feature vector. According to the technical solution provided by the present invention, the user portrait constructed based on big data can fully reflect the characteristics of the user, so that more refined services can be conveniently provided for the user according to the user portrait, and it overcomes the unary data that only relies on the user in the prior art, such as , The defects of low quality of user portraits caused by constructing user portraits such as age, gender, and income.

Fig. 2 shows a schematic flowchart of a method for constructing a user portrait based on big data according to another embodiment of the present invention. As shown in Figure 2, the method includes the following steps:

Step S200, acquiring the geographical location information of the user and the behavior information of the user operating the application program.

In step S201, the geographic location information is combined to generate a location information track.

User geographic location information can intuitively reflect user location changes, and is a means of analyzing user behavior. Analyzing a single user geographic location information may not find any abnormalities, but if multiple user geographic location information is combined for analysis, then It is easy to find abnormalities. For example, on xx, xx, xx, user A's user location information is Beijing in the morning, Shanghai in the noon, Sanya in the afternoon, and Beijing in the evening. For normal situations, a user It is unlikely to travel across four cities in one day. For this type of situation, if the problem cannot be found only based on the geographic location information of a single user, and the location information track is formed, it is easy to find the problem. Therefore, based on location Constructing user portraits from information trajectories can improve the accuracy of user portraits.

Specifically, after the geographic location information is obtained, the geographic location of the user is determined according to the geographic location information. Each geographic location is regarded as a location point, and the location points are connected with a straight line in chronological order to form a location information track. .

Step S202, acquiring information around the location point based on the location point on the location information track, and generating a sequence of location points.

The location information track is composed of location points. According to the location points on the location information track, query the information around the location point, and obtain the information around the location point, such as airport information, company information, shopping mall information, etc., according to the location point surrounding information and The geographic location information generates a sequence of location points.

Step S203, using the trained time recurrent neural network to perform sequence learning on the sequence of location points to obtain the feature vector of the user's geographic location corresponding to the user.

The time recursive neural network (LSTM) used in the embodiment of the present invention is obtained based on a large number of samples for training, and the generated position point sequence is input to the time recurrent neural network, and the time recursive neural network is used for the acquired position. The sequence of points is learned to generate the feature vector of the user's geographic location.

Step S204, obtaining the behavior track of the user operating the application program according to the behavior information.

The behavior information of the user operating the application program can intuitively reflect the situation of the user using the application program. According to the behavior information, it can analyze the user's focus on the application program, or analyze whether the user's operation is a kind of friendly behavior according to the behavior information.

For example, general applications require users to register an account before using the services provided by the application. Before the user registers, the corresponding disclaimer will be displayed to the registered user. Generally, friendly users will read the corresponding disclaimer. Know the corresponding precautions and fully understand the rights and interests of users. However, for some non-friendly users, they do not pay attention to the terms in the disclaimer, but hope to quickly complete the registration, enter the internal page of the application, and perform operations , such as borrowing, etc.

In many cases, analyzing a single user behavior information may not find any abnormalities, but if multiple user behavior information is combined and analyzed, it is easy to find abnormalities. Therefore, constructing user portraits based on behavior trajectories can improve the accuracy of user portraits .

Specifically, after the behavior information is acquired, the behavior information is connected with a straight line in chronological order to form a behavior track of the user operating the application program.

Step S205, generating a behavior sequence according to the behavior track of the user operating the application program.

Behavior trajectories connect behavior information with straight lines in chronological order. Therefore, behavior sequences can be generated according to the behavior trajectories of users operating application programs.

Step S206, using the trained time recurrent neural network to perform sequence learning on the behavior sequence to obtain the user behavior feature vector.

The time recursive neural network (LSTM) used in the embodiment of the present invention is obtained based on a large number of samples, and the generated behavior sequence is input to the time recurrent neural network, and the time recursive neural network is used for the acquired behavior sequence. Carry out learning and generate user behavior feature vectors.

Step S207, generating a location consistency feature vector according to the user's geographic location information.

The user geographic location information in this embodiment may include one or more of the following information: IP information, WIFI information, base station information, GPS information, but some information users can forge, for example, IP information, users can forge In this way, the user's geographical location determined based on the IP information will be inaccurate, and the inaccurate user's geographical location will lead to inaccurate user portraits constructed in the end, which will affect the use of user portraits to provide services to users. Therefore, after obtaining the user's geographic location information, it is also necessary to generate a location consistent feature vector based on the user's geographic location information. The consistency feature vector is used to represent the consistency of each geographic location and can be used to determine the user's geographic location. The eigenvectors are checked.

The present invention does not limit the execution order of steps S201-step S203, step S204-step S206, and step S207. Step S201-step S203 can be executed before or after step S204-step S206, step S207, and step S201 can also be executed at the same time , step S204, step S207.

Step S208 , performing verification processing on the feature vector of the user's geographic location according to the location consistency feature vector, to obtain a verified user's geographic location feature vector.

Specifically, the feature vector of the user's geographic location may be verified according to the location consistency feature vector, and the verified user's geographic location feature vector reflects the real user's geographic location.

Step S209, constructing a user portrait according to the verified user geographic location feature vector and user behavior feature vector.

The user portrait constructed by the user portrait construction method provided by the present invention can be applied to various platforms, enabling the platform to better understand users, thereby providing users with more refined services to improve user experience, for example , can be applied to game platforms, allowing game platforms to know the games that users are interested in, so as to push game advertisements to users in a targeted manner; it can also be applied to credit companies, so that the credit company can understand whether users have abnormal behaviors based on user portraits, Understand the user's credit records, and provide corresponding services to users based on the credit records, for example, do not lend to them, or determine the loan amount based on the records, etc. This is just an example and does not have any limiting effect.

In an optional embodiment of the present invention, after constructing the user portrait, the method further includes: training the user interest and hobby model based on the user portrait sample, inputting the user portrait of the user to be analyzed into the user interest and hobby model for analysis, and obtaining the user's interest Hobby.

Specifically, the user portrait constructed by using the user portrait construction method provided by the present invention is used as a sample for training, and the user's interest and hobby model with the input as the user portrait and the output as the user's hobbies and hobbies is obtained. When it is necessary to analyze the hobbies of a certain user, Utilize the user portrait construction method provided by the present invention to construct the user portrait of the user, then input the constructed user portrait into the trained user interest and hobby model, and use the user interest and hobby model to analyze and obtain the user's interests and hobbies. After the user's hobbies are analyzed, corresponding information may be pushed to the user according to the user's hobbies, for example, video messages, game messages, etc., which will not be listed here.

In an optional embodiment of the present invention, after constructing the user portrait, the method further includes: analyzing whether the user has abnormal behavior based on the user portrait.

For example, with the development of network technology, many lending applications have been launched. For example, xx IOU, users can borrow money in the application, and before lending to users, the company to which the application belongs needs to carry out risk assessment. The user's credit, etc., in the existing technology, is judged based on the credit records provided by the third party. However, risk control based on third-party data often faces many problems: 1. The data coverage of third-party data is limited. Many target users cannot be covered; 2. The cost of credit investigation is relatively high, and checking third-party data often requires user authorization, resulting in poor user experience. The user portrait construction method provided by the embodiment of the present invention is to construct a user portrait based on big data, make full use of the user's geographical location information and the behavior information of the user's operation of the application program, and avoid the user's falsified information, and the constructed user portrait is more realistic. After collecting user information, analyzing whether the user has abnormal behavior based on the user portrait can guide the credit company.

According to the method provided by the above-mentioned embodiments of the present invention, the geographical location information of the user and the behavior information of the user operating the application program are obtained, the geographic location information is combined and processed, the location information track is generated, and the surrounding information of the location point is obtained based on the location points on the location information track. , generate a sequence of location points, use the trained time recursive neural network to carry out sequence learning on the sequence of location points, obtain the user's corresponding geographical location feature vector, obtain the behavior trajectory of the user's operation application program according to the behavior information, and obtain the user's operation application program according to the user's operation application program. Behavior trajectory, generate behavior sequence, use the trained time recursive neural network to conduct sequence learning on the behavior sequence, obtain user behavior feature vector, generate location consistent feature vector according to user geographic location information, and classify user location according to location consistent feature vector The feature vector is verified, and a user profile is constructed according to the verified user's geographic location feature vector and user behavior feature vector. According to the technical solution provided by the present invention, the user portrait constructed based on big data can fully reflect the characteristics of the user, so that it can conveniently provide users with more refined services according to the user portrait, and determine whether the user has abnormal behavior according to the user portrait, etc. In the existing technology, only relying on the unary data of the user, such as age, gender, income, etc., to construct the user portrait leads to the defect of low quality of the user portrait.

Fig. 3 shows a schematic structural diagram of a device for constructing user portraits based on big data according to an embodiment of the present invention. As shown in FIG. 3 , the device includes: an acquisition module 300 , a first generation module 310 , a first learning module 320 , a second learning module 330 , and a user portrait construction module 340 .

The obtaining module 300 is adapted to obtain the geographical location information of the user and the behavior information of the user operating the application program.

The first generation module 310 is adapted to generate a location point sequence according to the user's geographic location information.

The first learning module 320 is adapted to use the first deep learning algorithm to learn the sequence of location points to obtain the feature vector of the user's geographic location corresponding to the user.

The second learning module 330 is adapted to use a second deep learning algorithm to learn behavior information to obtain user behavior feature vectors.

The user portrait construction module 340 is adapted to construct a user portrait based on the user geographic location feature vector and the user behavior feature vector.

According to the device provided by the above-mentioned embodiments of the present invention, the user's geographic location information and the behavior information of the user's operation of the application program are obtained, a location point sequence is generated according to the user's geographic location information, and the location point sequence is learned by using the first deep learning algorithm to obtain the user's corresponding The user's geographic location feature vector is used to learn the behavior information using the second deep learning algorithm to obtain the user behavior feature vector, and the user portrait is constructed based on the user's geographic location feature vector and the user behavior feature vector. According to the technical solution provided by the present invention, the user portrait constructed based on big data can fully reflect the characteristics of the user, so that more refined services can be conveniently provided for the user according to the user portrait, and it overcomes the unary data that only relies on the user in the prior art, such as , The defects of low quality of user portraits caused by constructing user portraits such as age, gender, and income.

Fig. 4 shows a schematic structural diagram of a device for constructing user portraits based on big data according to another embodiment of the present invention. As shown in FIG. 4 , the device includes: an acquisition module 400 , a first generation module 410 , a first learning module 420 , a second learning module 430 , a second generation module 440 , and a user portrait construction module 450 .

The obtaining module 400 is adapted to obtain the geographical location information of the user and the behavior information of the user operating the application program.

Wherein, the geographical location information of the user includes: IP information, WIFI information, base station information and/or GPS information. Behavior information includes: the user's operations on the application and the time spent on the page of the application.

The first generation module 410 further includes: a combination processing unit 411, which is suitable for combining the geographic location information to generate a location information track.

The first generation unit 412 is adapted to acquire information around the location point based on the location point on the location information track, and generate a sequence of location points;

The first learning module 420 is further adapted to: use the trained time recurrent neural network to perform sequence learning on the sequence of location points to obtain the feature vector of the user's geographic location corresponding to the user.

The second learning module 430 further includes: a behavior trace generating unit 431, adapted to obtain the behavior trace of the user operating the application program according to the behavior information.

The behavior sequence generation unit 432 is adapted to generate a behavior sequence according to the behavior track of the user operating the application program.

The second learning unit 433 is adapted to use the trained time recurrent neural network to perform sequence learning on the behavior sequence to obtain the user behavior feature vector.

The second generation module 440 is adapted to generate a location consistency feature vector according to the user's geographic location information.

The user portrait construction module 450 further includes: a verification processing unit 451, adapted to perform verification processing on the user geographic location feature vector according to the location consistency feature vector;

The user portrait constructing unit 452 is adapted to construct a user portrait according to the verified user geographic location feature vector and user behavior feature vector.

In an optional implementation manner of the present invention, the device further includes: a training module 460, adapted to train a user interest model based on a user portrait sample;

The interest and interest analysis module 470 is adapted to input the user profile of the user to be analyzed into the user interest and interest model for analysis to obtain the user's interest and interest.

In an optional implementation manner of the present invention, the device further includes: an abnormal behavior analysis module 480, adapted to analyze whether the user has abnormal behavior based on the user profile.

According to the device provided by the above-mentioned embodiments of the present invention, the user's geographic location information and the behavior information of the user's operation of the application program are obtained, the geographic location information is combined and processed, the location information track is generated, and the surrounding information of the location point is obtained based on the location point on the location information track. , generate a sequence of location points, use the trained time recursive neural network to carry out sequence learning on the sequence of location points, obtain the user's corresponding geographical location feature vector, obtain the behavior trajectory of the user's operation application program according to the behavior information, and obtain the user's operation application program according to the user's operation application program. Behavior trajectory, generate behavior sequence, use the trained time recursive neural network to conduct sequence learning on the behavior sequence, obtain user behavior feature vector, generate location consistent feature vector according to user geographic location information, and classify user location according to location consistent feature vector The feature vector is verified, and a user profile is constructed according to the verified user's geographic location feature vector and user behavior feature vector. According to the technical solution provided by the invention, the user portrait constructed based on big data can fully reflect the characteristics of the user, so that it can conveniently provide users with more refined services according to the user portrait, and determine whether the user has abnormal behavior according to the user portrait, etc. In the existing technology, only relying on the unary data of the user, such as age, gender, income, etc., to construct the user portrait leads to the defect of low quality of the user portrait.

The embodiment of the present application also provides a non-volatile computer storage medium, the computer storage medium stores at least one executable instruction, and the computer executable instruction can execute the user portrait based on big data in any of the above method embodiments Construction method.

FIG. 5 shows a schematic structural diagram of a computing device according to an embodiment of the present invention, and the specific embodiment of the present invention does not limit the specific implementation of the computing device.

As shown in FIG. 5 , the computing device may include: a processor (processor) 502 , a communication interface (Communications Interface) 504 , a memory (memory) 506 , and a communication bus 508 .

in:

The processor 502 , the communication interface 504 , and the memory 506 communicate with each other through the communication bus 508 .

The communication interface 504 is configured to communicate with network elements of other devices such as clients or other servers.

The processor 502 is configured to execute the program 510, specifically, may execute relevant steps in the above embodiment of the method for constructing a user portrait based on big data.

Specifically, the program 510 may include program codes including computer operation instructions.

The processor 502 may be a central processing unit CPU, or an ASIC (Application Specific Integrated Circuit), or one or more integrated circuits configured to implement the embodiments of the present invention. The one or more processors included in the computing device may be of the same type, such as one or more CPUs, or may be different types of processors, such as one or more CPUs and one or more ASICs.

The memory 506 is used for storing the program 510 . The memory 506 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.

The program 510 may be specifically configured to enable the processor 502 to execute the method for constructing a user portrait based on big data in any of the above method embodiments. For the specific implementation of each step in the program 510, please refer to the corresponding descriptions in the corresponding steps and units in the above-mentioned embodiment of user portrait construction based on big data, and details are not repeated here. Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working process of the above-described devices and modules can refer to the corresponding process description in the foregoing method embodiments, and details are not repeated here.

The algorithms and displays presented herein are not inherently related to any particular computer, virtual system, or other device. Various generic systems can also be used with the teachings based on this. The structure required to construct such a system is apparent from the above description. Furthermore, the present invention is not specific to any particular programming language. It should be understood that various programming languages can be used to implement the content of the present invention described herein, and the above description of specific languages is for disclosing the best mode of the present invention.

In the description provided herein, numerous specific details are set forth. However, it is understood that embodiments of the invention may be practiced without these specific details. In some instances, well-known methods, structures and techniques have not been shown in detail in order not to obscure the understanding of this description.

Similarly, it should be appreciated that in the foregoing description of exemplary embodiments of the invention, in order to streamline this disclosure and to facilitate an understanding of one or more of the various inventive aspects, various features of the invention are sometimes grouped together in a single embodiment, figure, or its description. This method of disclosure, however, is not to be interpreted as reflecting an intention that the claimed invention requires more features than are expressly recited in each claim. Rather, as the following claims reflect, inventive aspects lie in less than all features of a single foregoing disclosed embodiment. Thus, the claims following the Detailed Description are hereby expressly incorporated into this Detailed Description, with each claim standing on its own as a separate embodiment of this invention.

Those skilled in the art can understand that the modules in the device in the embodiment can be adaptively changed and arranged in one or more devices different from the embodiment. Modules or units or components in the embodiments may be combined into one module or unit or component, and furthermore may be divided into a plurality of sub-modules or sub-units or sub-assemblies. All features disclosed in this specification (including accompanying claims, abstract and drawings) and any method or method so disclosed may be used in any combination, except that at least some of such features and/or processes or units are mutually exclusive. All processes or units of equipment are combined. Each feature disclosed in this specification (including accompanying claims, abstract and drawings) may be replaced by alternative features serving the same, equivalent or similar purpose, unless expressly stated otherwise.

Furthermore, those skilled in the art will understand that although some embodiments described herein include some features included in other embodiments but not others, combinations of features from different embodiments are meant to be within the scope of the invention. and form different embodiments. For example, in the following claims, any of the claimed embodiments may be used in any combination.

The various component embodiments of the present invention may be implemented in hardware, or in software modules running on one or more processors, or in a combination thereof. Those skilled in the art should understand that a microprocessor or a digital signal processor (DSP) can be used in practice to implement some or all of some or all of the components in the user portrait construction device based on big data according to the embodiment of the present invention Function. The present invention can also be implemented as an apparatus or an apparatus program (for example, a computer program and a computer program product) for performing a part or all of the methods described herein. Such a program for realizing the present invention may be stored on a computer-readable medium, or may be in the form of one or more signals. Such a signal may be downloaded from an Internet site, or provided on a carrier signal, or provided in any other form.

It should be noted that the above-mentioned embodiments illustrate rather than limit the invention, and that those skilled in the art will be able to design alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses shall not be construed as limiting the claim. The word "comprising" does not exclude the presence of elements or steps not listed in a claim. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The invention can be implemented by means of hardware comprising several distinct elements, and by means of a suitably programmed computer. In a unit claim enumerating several means, several of these means can be embodied by one and the same item of hardware. The use of the words first, second, and third, etc. does not indicate any order. These words can be interpreted as names.

Claims (10)

1. The building method 1. a kind of user based on big data draws a portrait, including：

   Obtain user's geographical location information and user operates the behavioural information of application program；

   Position point sequence is generated according to user's geographical location information；

   The position point sequence is learnt using the first deep learning algorithm, it is special to obtain user geographical position corresponding to user Sign vector；

   The behavioural information is learnt using the second deep learning algorithm, obtains user behavior characteristic vector；

   Drawn a portrait based on user behavior characteristic vector structuring user's described in user's geographic location feature vector sum.
2. 2. the method according to claim 11, wherein, it is described to enter one according to user's geographical location information generation position point sequence Step includes：

   The geographical location information is combined processing, generation positional information track；

   Location point peripheral information is obtained based on the location point on positional information track, generates position point sequence；

   It is described that the position point sequence is learnt using the first deep learning algorithm, obtain user's geography position corresponding to user Characteristic vector is put to further comprise：

   Sequence Learning is carried out to the position point sequence using trained time recurrent neural network, obtained corresponding to user User's geographic location feature vector.
3. 3. method according to claim 1 or 2, wherein, it is described to utilize the second deep learning algorithm to the behavioural information Learnt, obtain user behavior characteristic vector and further comprise：

   The action trail of user's operation application program is obtained according to the behavioural information；

   The action trail of application program is operated according to user, generates behavior sequence；

   Sequence Learning is carried out to the behavior sequence using trained time recurrent neural network, obtains user behavior feature Vector.
4. 4. according to the method described in claim any one of 1-3, wherein, methods described also includes：According to user's geography position Confidence breath generation location consistency characteristic vector；

   It is described to be further comprised based on user's geographic location feature vector sum user behavior characteristic vector structuring user's portrait：

   Checking treatment is carried out to user's geographic location feature vector according to the location consistency characteristic vector；

   Drawn a portrait according to user's geographic location feature vector sum user behavior characteristic vector structuring user's after checking treatment.
5. 5. according to the method described in claim any one of 1-4, wherein, after structuring user's portrait, methods described also includes： Based on user portrait sample training user interest modeling hobbies；

   The user's portrait input for being analysed to user is analyzed to user interest modeling hobbies, obtains the interest love of the user It is good.
6. 6. according to the method described in claim any one of 1-4, wherein, after structuring user's portrait, methods described also includes： Drawn a portrait based on the user and analyze whether user has abnormal behaviour.
7. 7. according to the method described in claim any one of 1-6, wherein, user's geographical location information includes：IP information, WIFI information, base station information and/or GPS information.
8. The constructing apparatus 8. a kind of user based on big data draws a portrait, including：

   Acquisition module, the behavioural information of application program is operated suitable for obtaining user's geographical location information and user；

   First generation module, suitable for generating position point sequence according to user's geographical location information；

   First study module, suitable for learning using the first deep learning algorithm to the position point sequence, obtain user couple The user's geographic location feature vector answered；

   Second study module, suitable for learning using the second deep learning algorithm to the behavioural information, obtain user behavior Characteristic vector；

   User's portrait constructing module, suitable for based on user behavior characteristic vector structure described in user's geographic location feature vector sum Make user's portrait.
9. 9. a kind of computing device, including：Processor, memory, communication interface and communication bus, the processor, the storage Device and the communication interface complete mutual communication by the communication bus；

   The memory is used to deposit an at least executable instruction, and the executable instruction makes the computing device such as right will Ask the user based on big data any one of 1-7 to draw a portrait corresponding to building method to operate.
10. 10. a kind of computer-readable storage medium, an at least executable instruction, the executable instruction are stored with the storage medium Make user based on big data of the computing device as any one of claim 1-7 draw a portrait to grasp corresponding to building method Make.